Artrac Pinnacle 2022

Download as pdf or txt
Download as pdf or txt
You are on page 1of 183

VOL 21 ARTIFICIAL INTELLIGENCE AND THE FUTURE OF 2022

WARFARE: ROADMAP FOR THE INDIAN ARMY


PATRON PINNACLE - THE ARTRAC JOURNAL

Lt Gen SS Mahal, AVSM, VSM PINNACLE is one of the leading


professional journals of the Indian Army.
Published from the alma-mater of all
conceptual studies for future warfare, the
journal offers a vibrant forum to all serving and
VICE PATRON retired members of all three Services to share
military thoughts on contemporary security and
Lt Gen JS Sandhu, AVSM defence matters at the conceptual, directional
and functional levels.

The Indian Armed forces, like all other


EDITORIAL BOARD modern armies, have been influenced by
Information Warfare and Revolution in Military
Maj Gen VM Chandran, VSM - Chairman Affairs. The strategic thinkers and military
Brig Rajat Kumar - Chief Editor intellectuals need to explore the impact of these
Col Deepak Taluja - Editor emerging technologies on the doctrines and
concepts and visualize as to how it will affect
the method of waging wars in future.

In this era of ever changing battlefield


environment, it is incumbent on the part of the
servicemen as well as the military intelligentsia
to remain abreast with the developments. The
resultant changes in doctrines, strategies,
concepts, organisations and tactics need to be
evolved continuously. PINNACLE provides an
ideal forum to offer your valued comments on
how best we can adapt to the changes and
improve our military effectiveness.

PINNACLE is published every year.


Copyrights of all published articles rest with the
Editor, PINNACLE. No part of this publication
may be reproduced or transmitted in any form
without the written permission from the Editor.

The views expressed in the articles are


solely those of the authors in their individual
MISSION capacity. They do not represent the views of the
Headquarters, Army Training Command or the
“To offer a vibrant forum for serving Indian Armed Forces or any other agency of the
and retired members of the armed forces, Government of India.
civil services, police organisations,
members of government and academia to PINNACLE is registered with the
exchange the views and express mature Indian National Scientific Documentation
professional thoughts on contemporary Centre, with the International Standard Serial
national and international security matters, Number, 0972-4044, allotted to it. The title
art and science of warfare, military strategy, PINNACLE is also registered with the Office of
leadership, management and other topics the Registrar of Newspapers for India in terms
of vital national interest”. of the Proviso to Section 6 of the Press &
Registration of Books Act, 1867, vide their letter
No HPENG 00169/08/1/2001-TC dated
03 Jan 2002.
PINNACLE

Published By
Headquarters, Army Training Command, Shimla (HP), India

VOLUME 21 INDEX 2022

 Foreword (iii)
 From the Editor (iv)
 Theme for Next Issue (v)
ARTICLES
 Artificial Intelligence and its Brigadier (Dr) Navjot Singh 1-13
Exploitation in Future Warfare Bedi

 Strategy to Evolve Artificial Colonel Bikramjeet Singh 14-22


Intelligence in Tactical
Operations for the Indian Army

 Artificial Intelligence: Impact Colonel Harish Totade 23-32


on Warfighting and
Perspective Planning

 Artificial Intelligence in Cyber Miss Khyati Singh 33-42


Space: Towards Modern
Warfare

 Future of Artificial Intelligence: Colonel PS Mehta 43-47


Applications in the Indian Army
Communication Network

 Employment of Artificial Major Sharad Sagar Joshi 48-52


Intelligence to Enhance Efficacy
of Artillery

 Artificial Intelligence and Captain Shreya Sood 53-63


Machine Learning: The Way
Ahead for Futuristic Imagery
Interpretation

 Army Design Bureau- Colonel Amandeep Singh 64-71


Connecting Capabilities in Bains
Artificial Intelligence for New
Age Warfare

(i)
PINNACLE 2022
 Artificial Intelligence in Modern Captain SBS Bhullar 72-81
Warfare: Drone Detection using
Machine Learning

 Combat “AI-Effectiveness”: Role Colonel G Praveen, SM 82-90


of Artificial Intelligence in
Combat Effectiveness

 Leveraging Artificial Intelligence Brigadier S Balakrishnan 91-100


for Furtherance of Military and Lieutenant Colonel
Operations in the Indian Army Bikas Biswakarma

 Artificial Intelligence-Relevance Major Akhand Pratap 101-111


in Air Defence Battle

 Artificial Intelligence and Future Lieutenant Colonel Deepak 112-125


of Warfare: Roadmap for the Kumar Tiwari
Indian Army

 Artificial Intelligence and the Lieutenant Colonel Manu 126-134


Intelligentized Warfare Joseph Chacko

 Artificial Intelligence and the Brigadier Achal Dilip Kumar 135-149


Future of Indian Military
Logistics

 Legal and Ethical Aspects of Colonel Surendra Tanwar 150-155


Artificial Intelligence in Future
Wars

 Book Review: Artificial Lieutenant Colonel Manish 156-160


Intelligence and the Future of Rawat
Power: 5 Battlegrounds

 Book Review: AI by Design – A Colonel Jitender Kaushik 161-166


Plan for Living with Artificial
Intelligence

 PINNACLE Themes since 167


Inception

 Changing Character of War 168-171

 Guidelines for the Authors 172-173


(ii)
PINNACLE 2022
FOREWORD
Science and Warfare share a symbiotic relationship. While Geopolitical
considerations and Nature of War have proved enduring and timeless, technology has
th
shaped profound changes in the Character of War. In the 16 Century, science was used
to develop field fortifications and to calculate ballistic trajectories. It was however, in the
th
20 Century that technological inventions in the internal combustion engine and electronics
made their presence felt in unprecedented ways.
Warfare since the 1950s has been transformed by convergence of information and
computing prowess of the computer chip. Network Centric Warfare which facilitated the
convergence of sensors and communications often overwhelmed the decision makers with
information. Artificial Intelligence (AI), currently at nascent stage has already made
profound impact in administrative and logistics functions. AI also has the potential to
complement human cognitive functions, thereby, assisting commanders to take decisions
on the battlefield. In its evolved form, AI would reduce the human cognitive
Observe – Orient – Decide – Act (OODA) Loop into a more proactive and nonlinear
Anticipate – Act – Adapt paradigm, thereby transforming warfare.
Even as predicting the future is subject to probabilities, trends suggest that the
number of variables that could impact warfare in the future are likely to be greater than the
previous wars. Change in any one variable could create an exponential impact on the
outcome of the battle and even the war itself. Appreciating and monitoring the complex
interplay between large number of military and non-military variables are likely to transcend
beyond human cognitive abilities. In future wars, therefore, exploitation of AI to supplement
human cognitive abilities could represent the difference between the winning and losing
side.
Our Nation and the Armed Forces in particular have laid impetus in developing AI
enabled solutions. Institutions and framework have been established to incorporate this
niche technology in military operations to leverage the opportunities available to own
advantage in future battlefield. It gives me immense pleasure to release Pinnacle 2022
edition, themed „Artificial Intelligence and the Future of Warfare: Roadmap for the
Indian Army‟. The journal has articles and book reviews highlighting various facets of AI
and its applicability in warfare. I sincerely thank all authors for providing an insight on the
contemporary subject making it an interesting read.
JAI HIND

(SS Mahal)
Lt Gen
GOC-in-C, ARTRAC
(iii)
PINNACLE 2022
From the Editor

Dear Readers, the Editorial Team has the singular privilege of putting
forward this edition of Pinnacle with the theme „Artificial Intelligence and the
Future of Warfare: Roadmap for the Indian Army‟. Overwhelming response to
the theme is a clear indication of the relevance of the topic in the current scenario.
We are also grateful to all the readers who provided us with their valuable
feedback on our previous issue on „Employment of Offensive Cyber as a
Multiplication Tool for Defensive and Offensive Operations by the Indian
Army‟.

In the current issue, contributions have been made by a wide spectrum of


authors with varied service experience. The facets of „Artificial Intelligence and
the Future of Warfare: Roadmap for the Indian Army‟ that has been covered in
the Journal include application of Artificial Intelligence in operations to enhance the
combat efficiency of Indian Army. The journal highlights aspects of incorporating
Artificial Intelligence in cyber, communications, logistics, R&D and it‟s applicability
in future wars. The quality and content of each article received demanded that
they be published; however, due to constraints of space, the Editorial Team had to
contend with only 16 articles. Some of the articles have been edited to conform to
the size and to avoid repetition. We have also included two reviews of the books
„Artificial Intelligence and the Future of Power: 5 Battlegrounds by Mr Rajiv
Malhotra” and AI by Design – A Plan for Living with Artificial Intelligence by
Catriona Cambell‟ respectively.

Team Pinnacle expresses its sincere gratitude to our esteemed


subscribers, authors and readers for their keen interest in the Journal and look
forward for the contributions for our next issue based on the theme „Changing
Character of War’.

Happy Reading

- Editor

(iv)
PINNACLE 2022
Theme for PINNACLE 2023 Issue

CHANGING CHARACTER OF WAR


The Indian Army is a battle-hardened force renowned for its
professionalism. The rich experience gained in varied terrain and different
operations of war will undoubtedly be beneficial as the Indian Army prepares for
future conflicts. Leaders familiar with battlefield conditions are more likely to
appreciate, describe and lead their units and formations to success. However,
leaders would also need to relate their experiences within the context of a broad
understanding of the Nature and Character of War.

It is unlikely that experiences of the past conflicts will be fully applicable in


the future. Fog, friction, uncertainty and complexity have been intrinsic to the
Nature of War since times immemorial and are likely to remain so in the future.
However, the Character of War, which in essence is the form in which violence is
applied on the battlefield, has seldom been the same. Wars are a contest of wills
in a given social, cultural, political and economic backdrop; all of which continue to
evolve. Besides technology, structures and doctrines also influence the Character
of War. Leaders at all levels therefore, need to consider what experiences of the
past will be relevant in the future. This assessment will assist leaders in placing
their training/education as well as experiences in context to develop understanding
of the continuities and discontinuities of war.

Studying the evolving Character of War is important to ensure that the


Indian Army prepares for the right kind of war. Accordingly, Changing
„Character of War‟ has been identified as the theme for the next publication. The
publication would be divided in four themes with subthemes to ensure greater
granularity and insights into the Character of War. A note on Changing Character
of War has also been included along with guidelines for contributing authors for
reference at the end of this journal.

The professional journal being an unclassified venture and published on


the open domain, invites articles confined to the norms of unclassified domains of
the theme & sub-themes. I am extremely grateful to all the eminent authors for
sending in their contributions for the current edition on the theme „Artificial
Intelligence and the Future of Warfare: Roadmap for the Indian Army‟ and
would solicit similar support for the next issue.

- Chief Editor

(v)
PINNACLE 2022
DISCLAIMER

The views expressed and suggestions made in the articles are solely
of the author in his personal capacity and do not have any official
endorsement. Attributability of the contents lies purely with the author.

PINNACLE 2022
ARTIFICIAL INTELLIGENCE AND ITS EXPLOITATION
IN FUTURE WARFARE
- Brigadier (Dr) Navjot Singh Bedi

“It seems probable that once the machine thinking method had started, it
would not take long to outstrip our feeble powers… They would be able to
converse with each other to sharpen their wits. At some stage, therefore, we
should have to expect the machines to take control.”
- Alan Turing

Introduction

The above quote by Alan Turing is a stark reminder of what awaits us as


regards predicting Singularity. Despite that, among the newer generation
technological developments, Artificial Intelligence (AI), Quantum Computing and
Big Data Analytics hold greater promise for their applications in the Defence and
Security Forces. These largely computer science-based technologies are inspired
by the relentless research undertaken in the US and followed equally seriously in
China, Europe, Japan, Russia, and Israel. India has also made credible strides
and a Bengaluru-based DRDO Laboratory, Centre for Artificial Intelligence and
Robotics (CAIR), has developed several technologies for applications in Defence
Forces. The Indian Army has set up a centre for AI at the Military College of
Telecommunication Engineering (MCTE), Mhow for undertaking research in this
field. AI has the potential to revolutionise the way future wars will be fought. To
quote Ray Kurzweil, American inventor, and futurist, ―Artificial intelligence will
reach human levels by around 2029. Follow that out further to, say, 2045,
and we will have multiplied the intelligence – the human biological machine
intelligence of our civilization – a billion-fold.‖

On 18 November 2021, while delivering the keynote on India’s technology


evolution and revolution at the inaugural Sydney Dialogue, Hon’ble Prime Minister
Narendra Modi said that it is a "great honour" for India and that this indicates
India’s central role in the Indo-Pacific region and in the emerging digital world. He
said that ―We are in a time of change that happens once in an era. The digital age
is changing everything around us. It has redefined politics, economy and society. It
is raising new questions on sovereignty, governance, ethics, law, rights and
security. It is reshaping international competition, power and leadership". The
same is equally true for AI and the changes it will usher in the field of Defence.

Modern Warfare

Modern warfare refers to the concepts, methods and technologies that


have come into use during and after the Second World War and the Korean War.
The Gulf War (Operation Desert Storm) and the wars in Afghanistan, Iraq and
Russia-Ukraine are indicators of the shape of things to come. The concepts and
PINNACLE 2022 1
Artificial Intelligence and Its Exploitation in Future Warfare

methods have assumed complex forms of the 19th and early-20th-century


antecedents, largely due to the widespread use of highly advanced information
technology, and modern armies must modernize constantly to preserve their
battle-worthiness.1 Although total war was thought to be the form of international
conflict from the experience of the French Revolutionary Wars to the Second
World War, the term no longer describes warfare in which countries or nations use
all of their resources to destroy another country's or nation's organised ability to
engage in war. The practise of total war, which had been in use for over a century
as a form of war policy, has changed dramatically with greater awareness of
tactical, operational, and strategic battlefield.

Drivers of Change

The primary drivers of change in the way wars will be fought in times to
come will be the Economy, the Public Pressure/ Public Opinion, the Punitive Cost
of a Nuclear Strike, and the Impact of Emerging Technologies. In the present-day
context, due to the emphasis on economic development and the prohibitive costs
associated with wars, the likelihood of a full-scale war is extremely remote.
Moreover, due to the availability of nuclear weapons, full-scale war carries with it
the prospect of total annihilation, which is a no-win situation for all concerned. In
addition, no government relishes the prospect of its soldiers dying in combat and
being answerable to the public for the same. Lastly, the multitude of technological
developments that have taken place, including those that are on the drawing
board, have revolutionised the way wars will be fought.

As such, conflicts since WWII have by definition been "low intensity"


conflicts,2 typically in the form of proxy wars fought within local regional confines,
using what are now referred to as "conventional weapons," typically combined with
the use of asymmetric warfare tactics and applied use of intelligence and high
technology weapon platforms. The impact of AI in future warfare has been
analysed in subsequent paragraphs.

Introduction and Attributes of AI

Definition AI is an umbrella term for smart technologies that are aware of


and can learn from their environments, enabling them to subsequently take
autonomous action. Robotic process automation, machine learning, natural
language processing, and neural networks, all incorporate AI into their operations.
AI enables machines to respond autonomously to inputs from the external world,

1
International Congress Innovation &Technology XXI: Strategies & Policies Towards the XXI
Century, & Soares, O. D. D. (1997). Innovation and technology: Strategies and policies. Dordrecht:
Kluwer Academic.
2
Creveld, Martin Van (2000). "Technology & War I:To 1945". In Charles Townshend. The Oxford
History of Modern War. New York, USA: Oxford University Press. p. 206. ISBN 0-19-285373-2. And
Creveld, Martin Van. "Technology and War II:Postmodern War?". In Charles Townshend. The Oxford
History of Modern War. p. 349.
2
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

inputs that programmers do not directly control and therefore cannot always
anticipate3. AI is thus a collective term for computer systems that can sense their
environment, think, learn, and take action in response to what they are sensing.
Forms of AI in use today include, among others; digital assistants, chat bots and
machine learning.

Types of AI AI works in four ways, namely, Automated Intelligence


which deals with the automation of manual or cognitive and routine or non-routine
tasks; Assisted Intelligence which helps people to perform tasks faster and
better; Augmented Intelligence which helps people to make better decisions and
Autonomous Intelligence which leads to automating decision making processes
without human intervention.4 A brief description of a few technical terms usually
employed in AI literature often leads to a better understanding of its potential
applications.

Approaches to AI In an article titled ―Some Specific Tech Aspects of


Artificial Intelligence‖, André M. König, has described two vastly differing
approaches to AI5 namely: General and Narrow AI, also called Strong and Weak
AI. Strong AI is a hypothetical machine that exhibits behaviour at least as skilful
and flexible as humans, and there is a research programme to build such an
artificial general intelligence. A strong AI builds its own models based on raw input.
The principle behind Strong AI is that machines could be made to think or could
represent human minds in the future. Those machines will have the ability to
reason, think, and perform all functions that a human is capable of doing. Weak
AI, on the other hand, uses models of its problem domain assigned to it by
programmers. In Weak AI, machines can be made to act as if they are intelligent,
like a computer playing chess with a human player based on pre-programmed
moves by the humans

Machine Learning (ML) and Deep Learning (DL) ML uses algorithms to


parse data, learn from it, and then decide or predict something pertaining to the
data. The machine is trained using large amounts of data and algorithms that give
it the ability to learn how to perform the task. Computer vision is an application for
machine learning, though it still requires a great deal of hand-coding to get the job
done. Deep Learning (DL) is a branch of ML based on a set of algorithms that
attempt to model high-level abstractions in data. Simply put, one can have two
sets of neurons: ones that receive an input signal and ones that send an output
signal. When the input layer receives an input, it passes on a modified version of
the input to the next layer. In a deep network, there are many layers between the
input and output, allowing the algorithm to use multiple processing layers,

3
Chris Curran and Anand Rao; PWC Briefing: Artificial intelligence; http://usblogs.pwc.com/
emerging-technology/briefing-ai/.
4
ibid
5
Andre M Konig, “Some specific tech aspects of Artificial Intelligence”; https://www.linkedin.com
/pulse / some-specific-tech-aspects-artificial-intelligence.
3
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

composed of multiple linear and non-linear transformations. DL is part of a broader


family of machine learning methods based on learning representations of data.

Data Science Jim Gray imagined data science as a fourth paradigm of


science (empirical, theoretical, computational and now data-driven) and asserted
that everything about science is changing because of the impact of information
technology and the data deluge.

AI Development Trends

Messrs Rao, Voyles and Ramchandani have listed Deep Reinforcement


Learning (DRL) and Generative Adversarial Network (GAN) as emerging AI
trends in their paper on this subject6 and also indicated their likely future
applications. DRL involves interacting with the environment to solve business
problems. DRL has been used to learn gaming strategies. GAN is a type of
unsupervised deep learning system that is implemented as two competing neural
networks. One network, the generator, creates fake data that looks exactly like the
real data set. The second network, the discriminator, ingests real and synthetic
data.

Global AI Developments in Defence Applications7

Russia In 2017, Russian President Vladimir Putin declared that


whichever country becomes the leader in AI ―will become the ruler of the world.‖
Russia has been working on AI guided missiles that can decide to switch targets
mid-flight. Reportedly, there already exist completely autonomous AI operation
systems that provide the means for UAV clusters to fulfil missions autonomously,
share tasks between them, and interact. Russia believes that it is inevitable that
swarms of drones will one day fly over combat zones. Russia has been testing
several autonomous and semi-autonomous combat systems, such
as Kalashnikov's neural net combat module, with a machine gun, a camera
supported by AI that can possibly make its own targeting judgments without
human intervention. The Russian government has strongly rejected any ban on
Lethal Autonomous Weapons Systems (LAWS), suggesting that such a ban could
be ignored. The AI subsidiary of MTS,8 Russia’s largest telecom operator, has
launched a US$ 100 million venture capital fund for start-ups, to include upto
$20m in individual projects & start-up accelerator investing up to $100,000 per
project.

6
Anand Rao, Joseph Voyles and Pia Ramchandani;Top 10 artificial intelligence (AI) technology
trends for 2018; http://usblogs.pwc.com/emerging-technology/top-10-ai-tech-trends-for-2018/.
7
Artificial intelligence arms race; https://en.wikipedia.org/wiki/Artificial_intelligence_ arms_race.
8
https://developingtelecoms.com/telecom-business/telecom-investment-mergers/11904-russian-
operator-mts-launches-us-100m-ai-investment-fund.html
4
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

China China sees itself as a close competitor to the United States in AI.
The Chinese military intends to achieve an advantage through changing
paradigms in warfare with military innovation. The close ties between Silicon
Valley and China, and the open nature of the American research community, have
made the West's most advanced AI technology available to China. The Chinese
industry has numerous home-grown AI accomplishments of its own, such
as Baidu passing a notable Chinese-language speech recognition capability
benchmark in 2015. By 2030, China aims to make great strides in various AI
technologies like Big Data and Autonomous Intelligent Systems. As of 2017,
China’s roadmap aims to create a US$150 billion AI industry by 2030. China often
sources sensitive emerging technology such as drones and AI from private
start-up companies. The Japan Times reported in 2018, that private Chinese
investments in AI were under US$7 billion per year. In 2020, private investments in
AI from China amounted to almost US$ 9.9 billion in funding, ranking it second
after the USA. China's AI start-ups received nearly half of the total global
investment in AI in 2017. Of late, Chinese institutions have filed for nearly five
times as many AI patents as Americans have.

By 2025, China plans to achieve major breakthroughs in AI to reach a


leading level, with AI becoming a primary driver for China’s industrial
advancements and economic transformation. By then, China intends to become a
leading player in research and development while widely using AI in fields ranging
from manufacturing to medicine to national defence. China’s core AI industry is
likely to surpass 400 billion RMB (about US$ 59 billion), with AI-related fields
exceeding 5 trillion RMB (about US$ 740 billion). The Chinese leadership is
advocating an "innovation-driven" strategy for civilian and military development,
aiming to become the world’s "premier innovation centre" in AI by 2030.

China’s recent advances in Swarm Intelligence—which involves


autonomous cooperative behaviour among masses of distributed robots—have
been on prominent display in official media. In June 2017, China Electronics
Technology Group Corporation, a state-owned defence conglomerate,
successfully flight-tested a swarm of 119 drones—a new record. In a conflict, the
PLA could use swarms to target high-value U.S. weapons platforms, such as
aircraft carriers. In addition, China plans to achieve progress in the creation of
laws and regulations, as well as ethical norms and policies, along with the
establishment of mechanisms for AI safety assessment.

United States of America United States believes that the rapid


advances in AI will define the next generation of warfare. According to Deltek9,
identifiable federal spending on AI rose to nearly US $1 billion in 2020, up
50 percent from FY-2018, making it one of the fastest-growing emerging tech
investment areas. In 2020, private investments in AI in the US amounted to almost
$ 23.6 billion. Drones, motherships, protective exoskeletons, unmanned vessels,

9
https://crsreports.congress.gov/product/pdf/IF/IF11150.
5
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

and combat are some military-based AI applications in the US. A few other military
AI combat programmes of the US are the Sea Hunter autonomous warship, which
is designed to operate for extended periods at sea without a single crew member,
and to even guide itself in and out of port.

The USA government is spending billions of dollars preparing for the next
stage in warfare that it believes will be defined by advances in AI. Concepts like
motherships of drones releasing little baby drones from the air and the sea,
infantrymen and women sporting exoskeletons and wearable electronics loaded
up with combat apps, and lone mission commanders directing swarms of
unmanned vessels to carry out operations are already being tested at MIT’s
Computer Science and AI Laboratory.

A new National Science and Technology Council (NSTC) Subcommittee on


Machine Learning and Artificial Intelligence was formed on 03 May 2016 to help
coordinate federal activity in AI. On June 15, 2016, NSTC directed the
subcommittee on Networking and Information Technology Research and
Development (NITRD) to create a National Artificial Intelligence Research and
Development Strategic Plan. A NITRD Task Force on Artificial Intelligence was
then formed to define the federal strategic priorities for AI R&D, with particular
attention to areas of national security that private industry is unlikely to address. In
the last few decades, one of the largest sources of funding for AI research came
from the Defence Advanced Research Project Agency (DARPA).

Israel Most remarkably, Israel has seen investments in AI start-ups go


from 5% of all equity invested in start-ups in 2011 to 25% by mid-2018. The Israel
Innovation Authority (IIA) has launched a national programme to promote AI and
data science at a total cost of 5.26 billion shekels (US$ 1.63 billion). The goals of
the programme are to promote high-level research to create an ecosystem that will
enable the continued development and competitiveness of industry, academia,
and the public sector. It also aims to develop critical applications for Israel's
security and to assimilate the use of AI in industry, services, and government
ministries. Israel has designed Harpy anti-radar fire and forget drone that is to be
launched by ground troops and autonomously fly over an area to find and destroy
radar that fits pre-determined criteria.

Military Applications of AI10

Adopting AI tools in defence enhances the processing and utilization of


data which in turn improves the speed of decision-making on the battlefield. Below
are few applications of AI in the military:-

 Computational Military Reasoning Computational military


reasoning solves military problems that humans face and focuses on

10
https://acadgild.com/blog/applications-of-ai-in-military.
6
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

making the right battlefield decisions. It analyses the area of conflict and
acts on the data it receives from the set of orders (known as the "Course of
Action" or COA). In controlled settings like military labs, this application can
exploit the weaknesses of the enemy successfully.

 Intelligent and Autonomous Unmanned Weapon Systems


Unmanned vehicles that are used by armed forces, as well as the most
sophisticated weapons and robotics, are some of the direct uses of AI.
These weapon systems have the requisite intelligence to observe, pursue
and destroy enemy targets distinctly.

 AI-Enabled Information Processing and Intelligence Analysis


Deep learning algorithms can be used to effectively process sensor data
and raw intelligence that has been collected from satellite imagery. This
information can aid in making accurate decisions during military operations.

 Communications and Computers Communication is the core of


all military activities, providing improved visibility. The non-availability of
real time information can hamper the decision-making ability of soldiers
fighting the war. In future, the success on the battlefield will depend on
maintenance of network connectivity and management of information,
especially using agile cognitive networks, from a large variety of sources,
which will be considerably enhanced due to AI.

 Cyber Defence and Cyber Warfare Cyber-attacks can be


detected and curbed through pattern matching, statistical analysis,
machine learning and big data analysis. Offensive cyber operations can
unleash large scale destruction of cyber networks in a matter of minutes.
Considering the speed at which AI works, it serves as a counter-intuitive
force against malicious cyber threats. Generative Adversarial Networks
(GAN) are likely to be employed in cyber detection applications in future.

 Electronic Warfare (EW) EW employs the electromagnetic


spectrum (that includes radio waves, infrared signals, or radar) to sense,
protect, and communicate data. It can also be used to block adversaries
from using electronic equipment. In India, Adaptive Radar
Countermeasures (ARC) and Behavioral Learning for Adaptive Electronic
Warfare (BLADE) are direct applications of AI. ARC enables airborne EW
systems to automatically generate effective spontaneous countermeasures
against unknown radars in real time. This is done intelligently in the
presence of other signals, which could be either hostile or neutral. On the
other hand, the BLADE program develops machine learning algorithms to
rapidly detect new radio threats and dynamically generate new
countermeasures that are similar to ARC even in the most tactical
environments.

7
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

 Pattern Recognition Pattern recognition is the area of research


that studies the operation and design of systems that identify patterns in
data. When a program makes observations of some kind, it is often
programmed to compare what it observes as a pattern, e.g., face,
fingerprint, or handwriting recognition. Important application areas are
image analysis, character recognition, speech analysis, man-machine
diagnostics, and personnel identification, all of which have applications in
the military.

 Bioinformatics Bioinformatics is the application of computer


technology for the management of biological information. AI provides
several powerful algorithms and techniques for solving important problems
in bioinformatics. Approaches like Neural Networks, Hidden Markov
Models, Bayesian Networks and Kernel Methods are ideal for areas with
more data but very less theory. The goal in applying AI to bioinformatics is
to extract useful information from the wealth of available data by building
good probabilistic models.

 Data Mining An AI-powered tool that can discover useful


information within a database that can then be used to extract patterns
from data. Data mining is seen as an increasingly important tool by modern
businesses to transform data into business intelligence, giving an
informational advantage. It is currently used in a wide range of profiling
practices, such as marketing, surveillance, fraud detection, and scientific
discovery. Profiling and surveillance have applications, especially in CI/CT
operations.

 Expert Systems An expert system is a computer program that


represents reason with knowledge of some specialist subject with the view
of solving problems or rendering advice. It is essentially a
knowledge-based system, having several modules, sub-modules, and
sections. The knowledge-based applications of AI have enhanced
productivity in almost all fields, such as business, science, engineering,
and the military. Today’s expert systems clients can choose from dozens of
commercial software packages with easy-to-use interfaces. Diagnosis and
troubleshooting explain the development and testing of a condition-
monitoring sub-module of an integrated plant maintenance management
application based on AI techniques.

8
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

 Robotic Surgery11,12&13 This has the potential to relive


surgeons to perform other lifesaving tasks/supervise robotic surgery and
has immense potential in posts at remote/inaccessible areas.

 Powered Exoskeleton14 AI coupled with robotics can facilitate


development of Future Force Warrior (like Iron-man). This will not only aid
heavy lifting but will also provide a solution for paralysis/muscle related
diseases, warfare, construction, LOPES (exoskeleton), Re Walk, Human
Universal Load Carrier.

 Body Implants, Prosthesis, Regenerative Medicine Brain


implant, retinal implant, prosthetics, rejuvenated soldiers & life extension
may soon be possible due to AI. This will greatly augment the combat
potential and staying power of the armed forces.

 Self-Reconfiguring Modular Robot (SCMR) AI will enable


SCMR to be a reality which may change the way we make many physical
structures and machines.

 Swarm Robotics15 AI enabled swarm intelligence, autonomous


robotics, nanorobotics, particle swarm optimization, multi-agent systems,
behaviour-based robotics have the potential to ensure autonomous
cooperative behaviour among masses of distributed robots. In a conflict,
swarms of drones can be used to target high-value weapons platforms of
the adversary, such as aircraft carriers, as well as to confuse and evade
enemy radar.

 Cloak of Invisibility16&17 AI has the potential to provide


camouflage cloaking microscope tips at optical frequencies, thus
addressing one of the challenges that has perplexed armies since time
immemorial.

11
"Doctors grapple with the value of robotic surgery". Houston Chronicle 16 September 2011.
Retrieved 24 December 2011.
12
"Robotic surgery making inroads in many medical procedures." The Jakarta Post 8 March 2011.
Retrieved 24 December 2011.
13
"Doctors Perform First Fully Robotic Surgery". PC World. 21 October 2010. Retrieved
24 December 2011.
14
Christopher Mims (2009). "Exoskeletons Give New Life to Legs". Scientific American. Retrieved
21 April 2009.
15
"Riders on a swarm". The Economist. 12 August 2010. Retrieved 21 April 2011.
16
Rachel Kaufman (28 January 2011). "New Invisibility Cloak Closer to Working "Magic"". National
Geographic News. Retrieved 4 February 2011.
17
"Breakthrough in bid to create 'invisibility cloak' as 3D object is made to vanish for first time". Daily
Mail. 26 January 2012. Retrieved 3 March 2012.
9
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

 Immersive Virtual Reality AI can also provide an artificial


environment where the user feels just as immersed as they usually feel in
consensus reality. This will add a realistic feel to the conduct of war games
and simulations and may end up economising expenditure on formation
exercises.

 Social Media based IW With social media sites overtaking


TV as a source for news for young people and news organisations
increasingly reliant on social media platforms for generating distribution,
major publishers now use AI to post stories and generate higher volumes
of traffic. The same rationale can be used by the ADG PI in furthering a
favourable narrative.

 Logistics Various models of operations research have been


employed in effective management of logistics operations. AI has great
potential in assisting in planning and keeping the supply chain effective and
efficient. In 1991, US forces deployed a Dynamic Analysis and Re-planning
Tool (DART), using AI planning techniques, during the Gulf War to
undertake automated logistics planning and scheduling for transportation,
which allowed a plan to be generated in hours rather than weeks.

Tactical Applications of AI18 The advent of AI could fundamentally


change the character of war, resulting in transformation from today’s ―informatized‖
ways of warfare to future ―intelligentized‖ warfare, in which AI will be critical to
military power. Few tactical applications of AI are as under:-

 Mine sweeping drone bots that use feature maps to analyse and
identify mines, to deactivate them or carry them away.

 Enemy segmentation detection - identify enemy tanks in different


situations.

 Combat simulations using virtual reality, can help train soldiers for a
more realistic battlefield environment.

 Combat helmets with visors help analyse battlefield environment


and provides wide vision.

 Analysing location of missiles fired, akin to satellite defence grid


systems to shoot down the correct missile.

18
Military Applications Of Artificial Intelligence; March 17,2018 By Deepak Kumar Gupta
(http://www.claws.in/1878/military-applications-of-artificial-intelligence-deepak-kumar-gupta.html).
10
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

 Screenless Display19 (Virtual Retinal Display, Bionic Contact


Lens), Eye Tap Augmented reality, virtual reality, Eye Tap could allow
user to reference the blueprints like in a construction yard, in a 3D manner,
Head-mounted display, Head-up display, & adaptive optics for the next
generation soldier.

 Electronic Nose20&21 Detecting spoiled food, chemical weapons


and explosives and detection of contagious pandemic affected people.

 Memristor22&23 AI enables the development of smaller, faster,


lower power-consuming analogue electronics, which has immense
potential in equipment used at the tactical level.

 Machine Vision24 Biometrics, controlling processes (driverless


car, automated guided vehicle), detecting events (visual surveillance),
interaction (non-human-computer interaction), and robotic vision.

Niti Ayog Roadmap The national AI task force at NITI Aayog of India has
also set up a road map for the AI development employing intelligent machines
enabling high-level cognitive processes like thinking, perceiving, learning, problem
solving and decision making, coupled with advances in data collection and
aggregation, analytics and computer processing power. NITI Aayog has decided to
focus on five sectors that are envisioned to benefit the most from AI in solving
societal needs:-

 Healthcare: increased access and affordability of quality healthcare,

 Agriculture: enhanced farmers’ income, increased farm productivity


& reduction of wastage,

 Education: improved access and quality of education,

 Smart Cities and Infrastructure: efficient and connectivity for the


burgeoning urban population and

19
"Google 'to unveil' hi-tech Google Glasses that put a screen of information over the world." Daily
Mail 20 December 2011 Retrieved 22 December 2011.
20
"Tuberculosis breakthrough as scientists get funds for 'electronic nose'". The Guardian.
07 November 2011. Retrieved 4 December 2011.
21
"Now, a mobile phone that can smell". The Times of India. 7 November 2011. Retrieved
04 December 2011.
22
"Remapping Computer Circuitry to Avert Impending Bottlenecks". The New York Times.
28 February 2011. Retrieved 27 April 2011.
23
"Memristor revolution backed by HP". BBC News. 2 September 2010. Retrieved 27 April 2011.
24
"The big plan to build a brain". The Telegraph. 21 June 2011. Retrieved 18 November 2011.
11
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

 Smart Mobility and Transportation: smarter and safer modes of


transportation and better traffic and congestion problems.

AI Commercial Infrastructure25 Key trends of AI commercial


infrastructure are:-

 AI Industry in India was around US$180 Million annually in


revenues as on 2018. India's AI market is expected to reach $7.8 billion
by 2025 growing at a CAGR of 20.2%, as per an International Data
Corporation (IDC) report. AI software segment would dominate the market
and grow from $2.8 billion in 2020 at a CAGR of 18.1% by the end of 2025.

 There are approx one lakh AI personnel working in India across


enterprises and sectors – this represents a 20% jump in personnel from
2021 (91000 AI personnel) & from 29,000 in 2018. According to a 2018
report the need for AI skills has more than doubled in the past three years,
with job postings going up by a whopping 119%. The average work
experience of AI professionals in India is 7.6 years.

 University of Mumbai, BITS (Pilani), IITs (Kharagpur, Delhi,


Mumbai, Kanpur, Roorkee), University of Pune & Delhi are the top
universities that are undertaking professional AI graduate/ post graduate
courses.

Conclusion

AI will eventually manifest in all dimensions of warfare. A comprehensive


long-term vision will help formulate a mission oriented long-term policy for critical
strategic AI technologies. Since the locus of innovation has shifted to the private
sector, the government should facilitate to create closer public-private
partnerships, with an emphasis on "Make in India". Human resource needs to be
trained and cultivated for military applications of AI.

Home to the world’s third largest start-up ecosystem, elite science and
technology institutions like the IITs, robust and ubiquitous digital infrastructure, and
millions of newly minted STEM graduates every year, India is well-positioned to
become a global leader in the development of artificial intelligence. Industry
analysts predict that AI could add up to $957 billion to India’s economy by 2035.

In the spirit of Sabka Saath Sabka Vikas, Prime Minister Shri Narendra
Modi plans to leverage AI for inclusive development, representing the country's ―AI
for All' strategy. Directed by the PM’s vision, India will soon stand out in the
international community not just as a leader in AI, but also as a model to show the

25
Military Applications Of Artificial Intelligence; March 17, 2018 By Deepak Kumar Gupta,
(http://www.claws.in/1878/military-applications-of-artificial-intelligence-deepak-kumar-gupta.html).
12
PINNACLE 2022
Artificial Intelligence and Its Exploitation in Future Warfare

world how to responsibly direct AI for social empowerment. In his words “India is
one of the leading nations in Artificial Intelligence and Machine Learning,
especially in human-centred and ethical use of artificial intelligence. We are
developing strong capabilities in cloud platforms and cloud computing. This
is key to resilience and digital sovereignty."

AI has the potential to transform national security. It is also seen that


AI is essentially a dual-use technology. While it can fuel technology-driven
economic growth, it also has the potential to provide military asymmetry.
India has strong IT industry and huge talent pool of engineers which are
advantages which need to be leveraged.

Bibliography

1. Niti Ayog National Strategy for AI: Discussion-Paper; June 2018.

2. Artificial Intelligence And The Future Of Defense Strategic Implications For SmallAnd
Medium-Sized Force Providers; The Hague Centre for Strategic Studies-2017.

3. Warfighting 2018 & Beyond; CENJOWS Aug 2013, by Col Navjot Singh.

4. Artificial Intelligence and the Future of Warfare; Research Paper M. L. Cummings


International Security Department and US and the Americas Programme January 2017.

5. Artificial Intelligence in Military Operations: Technology & Ethics Indian Perspective; Lt Gen
RS Panwar, AVSM, SM, VSM, PhD (Retd); January 2019 - March 2019.

6. Artificial Intelligence In War: Human Judgment as an Organizational Strength and a


Strategic Liability; Avi Goldfarb and Jon Lindsay November 2020; The Brookings Institution
publication.

7. Future Warfare and Artificial Intelligence- The Visible Path; Gp Capt Atul Pant; IDSA
Occasional Paper No. 49; August 2018.

8. Military Applications of Artificial Intelligence; Ethical Concerns in an Uncertain World; by


Forrest E. Morgan, Benjamin Boudreaux, Andrew J. Lohn, Mark Ashby, Christian Curriden, Kelly
Klima, Derek Grossman; RAND Corporation, Santa Monica; 2020.

9. International law & the military use of unmanned maritime systems Michael N. Schmitt &
David S. Goddard; International Review of the Red Cross (2016), 98 (2), 567–592. War & security at
sea

10. Final report of National Security Commission on AI- Mar 2021 Report; Eric Schmidt –Chair
& Robert Work Vice Chair.

11. Scope for Use of AI in Security: Opportunities and Challenges; 31st IISSM Annual Global
Conclave- 2021; Maj Gen PK Mallick, VSM (Retd); 16 December, 2021.

12. Artificial Intelligence (AI) and Its Applications For Defence; CENJOWS/ Indian Def Conclave
Seminar; Brig (Dr) Navjot Singh Bedi; Dec 2018.

13
PINNACLE 2022
STRATEGY TO EVOLVE ARTIFICIAL INTELLIGENCE IN
TACTICAL OPERATIONS FOR THE INDIAN ARMY
- Colonel Bikramjeet Singh

“We can evade reality, but we cannot evade the consequences of evading
reality”
- Ayn Rand

Introduction

Despite an early start in the year 1950, the real development in the field of
Artificial Intelligence (AI) started only around 2010 due to the confluence of three
major enablers in the field: availability of big sources of data; massive
improvement in computing and processing power; and most importantly, the
massive improvement in machine learning science. AI today is being heavily
invested in and researched upon in the scientific and commercial world. It is
already showing up for nascent commercial usage in many gadgets and devices,
search assistance, requirement prediction, data analysis and validation, modelling
and simulation, linguistics, psychology etc.

AI now is also termed the Industrial Revolution 4.0 is bringing significant


transformations in the commercial as well as military domains. In December 2021,
the Chief of Army Staff, General M. M. Naravane, said, "Future wars will be fought
with inventive methods and AI." As our needs converge toward the use of AI in the
commercial domain, we must also pay attention to the use of AI in military training,
warfare, and wargaming. The potential of AI lies in how effectively one can
combine human interpretation, data, and algorithms to produce results, enhance
the effectiveness of training and operational readiness of the Indian Army. This
article will try to analyse the key facets of the same.

Use of AI During Tactical Operations in Modern Day Warfare

As we grow in the domain of AI, there is a requirement for developing,


integrating, and enabling advanced AI-based systems that will help the military or
the armed forces in combat at the tactical level. While it is known that cloud-based
AI systems can be used at operational or theatre level, these systems lack
proximity to the tactical edge. Several advanced states or countries have
conducted research to make effective use of AI in the tactical domain, and as a
result, certain applications where such systems can be used have been
discovered. Such applications of AI in tactical operations are listed in the
succeeding paragraphs.

 Autonomous Machines The use of robotics for taking complex


decisions, which are human-like, is mostly done using autonomous
machines. The development of autonomous vehicles has been progressed

PINNACLE 2022 14
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

on this concept. Various applications include obstacle recognition,


navigation and explosive decanting, etc.

 Mules Robots which can work along with armed forces in the
forward areas may be a far-fetched dream as of now. However, in some
countries, research has already gone into the provision of logistic-based
robots which can perform tasks like evacuation of casualties, delivering
heavy loads or acting as mules (e.g. the BigDog project of Defence
Advanced Research Projects Agency (DARPA)). Defence Research and
Development Organisation (DRDO) is also looking forward to develop a
Multi Agent Robotics Framework (MARF) which can be used for
surveillance and casualty evacuation in Counter Terrorism (CT) operations.

 Swarming of Drones AI-based cooperative behaviour, or


swarming, is a unique subset of autonomous vehicle development, with
concepts ranging from large formations of low-cost drones designed to
overwhelm defensive systems to small squadrons of drones that
collaborate to provide electronic attack, fire support, and localised
navigation and communication nets for ground-troop formations. AI
technology may also be adapted for defending harbours, hunting
submarines, or scouting in front of a formation of larger navy ships and
tanks.

 Logistics The validation and audit of transportation decisions and


their optimal utilisation for the logistics requirements of armed forces can
be done through AI-based programs. Similarly, meeting logistics
requirements, including movement of troops or material, is a herculean
task, especially in high-altitude areas and difficult terrain. This task, if
automated with AI, will reduce a huge amount of effort at the operational
level. Knowledge Resource and Intelligent Decision Analysis (KRIDA), an
application which aims to facilitate military logistic movements at large
scales using extensive knowledge base and data editing systems, is one of
the major projects taken up by Centre for Artificial Intelligence and Robotics
(CAIR) & DRDO.

Use of AI in Defence and Global Military Conflicts

 Armenia and Azerbaijan War In 2020, the news channels were


bloated with the conflict between Azerbaijan and Armenia, not only
because the conflict was over the Nagorno-Karabakh region, but for
Azerbaijan’s overwhelming and decisive defeat of Armenia by using AI-
based complex military hardware. The country demonstrated how the use
of technology such as armed drones can enable minor adversaries to
outmanoeuvre a well-equipped enemy on the battlefield. During the
conflict, Armenian soldiers apparently anticipated fighting a war as they
fought in 1994, using Russian-built T-72 tanks covered by an intimidating
15
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

network of extremely effective S-300 high-alt air defence missile systems.


However, the Armenian forces were outright annihilated by the Azerbaijani
forces, which were already prepared for the war of the future.

 Russia Drone swarming of the combat zone is inevitable in future


wars. According to reports, Russia has developed AI systems capable of
performing autonomous or shared missions via drone clusters. They are on
the path of achieving AI guided missiles, which have the capability of
midflight target switching. Semi-and fully automated combat systems, such
as Kalashnikov's AI neural net combat module integrated with a machine
gun and a camera for target recognition, are also being tested. Russia is
following the aggressive policy of no ban on the development of lethal
autonomous weapons.

 China Deep ties between China and Silicon Valley technology


firms have allowed Chinese home-grown technology firms access to the
most advanced AI technology available. Number of AI patents held by
Chinese institutions is five times more than what is held by American. Also,
drone and AI technology are sourced from private firms by China. In 2015,
Chinese homegrown search engine Baidu was able to crack the speech
recognition capability benchmark. By 2030, China is on the path of creating
a $150 billion AI industry. As per a 2018 report by the Japan Times,
Chinese private annual investment in AI is close to $7 billion.

 Israel Fire and Forget drone, "Harpy," developed by Israel, has


the capability of autonomous flying over an area to find and destroy enemy
radars that fits a predetermined criterion.

 United States of America The US is working with the commercial


sector to harness AI at the tactical level. In 2021, US private investment in
AI was close to $100 billion. IBM is analysing the United States Army’s
Logistics Support Activity (LOGSA) transportation of spare parts. AI combat
programmers such as sea hunters’ autonomous warships have been
designed which enable crew-less operations at sea for extended periods
with the ability to guide themselves in and out of ports.

 Russia-Ukraine War Despite being quantitatively weaker, Ukraine


has put up tough resistance against Russian invasion with the use of
advanced drones facilitated by Elon Musk’s Starlink satellite system, which
is enabling reconnaissance missions by providing uninterrupted internet
connectivity. Turkey’s AI-based TB2 Bayraktar drones have been put to
effective use by the Ukrainian force. In retaliation, the Russian Army is also
exploiting Orion combat drone system which was also used in the Syrian
conflict in 2015.

16
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

Status of AI in Indian Armed Forces

AI can be used anywhere and


everywhere; however what one needs to
understand is that the effort required for
developing the same is immense. The
figure below shows us two solution spaces
for AI, i.e., our defence requirements and
solution space available for development.
The exact requirement lies between the
intersection of both the spaces, as depicted
in Figure 1. This, however, is an abstract
issue as effort, time, and cost vs. the need
are certain issues that need to be taken into consideration while developing or
undertaking defense-related projects in Robotics and IA. AI has a typical ‘dual-use’
description. The situation is not about humans vs. machines but about interface
between humans and machines to make better decisions.

Lethal Autonomous Weapon Systems (LAWs) No country has


currently deployed fully autonomous weapon systems. However, all global leaders
recognised a need to address issues concerning the morality of these weapons
and their status under international humanitarian law. The United Nations (UN)
has been arguing for a preemptive ban on the development of such weapons.
However, countries like the United States have argued that autonomous weapons
have benefits, including reducing casualties and improving efficiency in defensive
capabilities.

AI Roadmap for India The AI roadmap by the Indian government was


published in the year 2018 by the AI task force, which was constituted by the
Ministry of Commerce and Industry and the national strategy for artificial
intelligence under the stewardship of Niti Aayog. While acknowledging the need to
promote AI, governments at different levels, along with their various
instrumentalities, should adopt proactive measures to accelerate AI adoption in
various processes. However, the report of the task force on AI does recognise AI
as a major force multiplier and suggests development military applications. The
report has clearly given and identified that the cost of technology and system
management capability are major challenges towards developing military
applications for AI.

Doctrines Though the doctrines for integrated warfare exist with the
Indian armed forces, stringent policies exist for evolving any integrated procedures
for data infusion. This will result in inconsistent logic in algorithms, which
eventually forms the backbone of any AI-based systems. Also, the rapid changes
in Tactics, Techniques and Procedures (TTPs) of the Armed Forces is a major

17
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

aspect that needs to be kept in mind to avoid incorrect decision making by AI


systems.

Foreign Acquisition The acquisition process for any AI-related project


will include the sharing of database and TTPs, which will have security related
issues. Dummy data will have impact accuracy of AI solutions. Moreover, present
acquisition processes are not agile enough for fast-paced development in AI.

Hardware and Microelectronics The power of AI comes from the


hardware and the microelectronics chips being used in it. Thus, we need to
fabricate microchips to ensure secure and cost-effective AI-based systems that will
shape the future battlefield.

Strategy for Evolving AI based Defence Capabilities for the Indian Army

AI is a technology that will provide the Indian Army with greater advantages
than any other technology. Other technologies such as big data, cryptography, and
cyber aspects can also be proliferated to the next level using AI. However, to fully
exploit AI, there is a requirement to go ahead with certain requirements shown in
Figure 2 that will come a long way in the future. Some of them are highlighted in
the succeeding paragraphs.

 Data Storage AI data storage and its scaling is one of the biggest
contemplations as the volume of data for AI keeps on growing. As
organisations concoct AI stratagems and build the essential structure,
storage must be of the utmost precedence. Storage will include ensuring
the appropriate storage capacity and consistency to deal with the
enormous data required for AI. Typically, the storage may depend on
factors which include the level of AI, real-time decisions, and the source of
data. As databases cultivate with time, organisations need to monitor
capacity and plan for expansion as needed. Moreover, for military data, this
data needs to be secured to prevent any misuse.

 Data Availability and Integrity AI algorithms need to be trained


on large datasets. If a technologically capable adversary is aware of the
dataset used by AI, they will have an advantage in deciphering ways to
defeat it. In case the enemy inserts false data into our systems, the AI will
learn the untrue version of reality that the adversary desires. Under these
circumstances, it will be very dangerous, especially because the victim will
not realise that their own AI dataset is incorrect and may continue trusting
the incorrect results. Moreover, the availability of data for training the AI-
based tactical operations is also questionable.

18
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

Data Storage

Civil-Defence
Prepare Data
Cooperation

Processing
Skill Set Strategy
Data

Data Networking
Management Infrastructure

Processing
Tools

Figure 2: Strategy for Evolving AI based Defence Capabilities

 Networking infrastructure One of the key components of the AI


infrastructure is its networking infrastructure. AI learning algorithms are
highly dependent on communications and enterprise networks will be
essential to keep pace with the demand as AI efforts expand. AI systems
will require high-bandwidth and creative architectures. The Network for
Spectrum can be a stepping stone towards the primary needs for AI
networking infrastructure.

 Processing and Preparing AI Data Not only is there a need to


plan storage, movement and processing of data for AI, but we need to
choose how to prepare the data for use in AI applications. This is one of
the critical steps for cleansing the AI data and is sometimes also called
data scrubbing. This process includes updating or removing data from a
database that is inaccurate, incomplete, improperly formatted or duplicated.
Thus, adequate infrastructure in terms of manpower, hardware and tools
must be procured for preparing the AI data.

 Data Management and Governance Another important factor is


data access. An organisation must have proper mechanisms in place to
deliver data in a secure and efficient manner to the required users. A data
management strategy is needed to ensure that human users and machines
have easy and fast access to data via wireless networks. Access also
raises several privacy and security issues, so data access controls are
important. We need to look at technologies such as identity and access

19
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

management and data encryption tools as part of data management and


governance strategies.

 Skillset Training and skills development are vital for any


endeavour. AI will need data analysts, data scientists, developers, cyber
security experts, network engineers, and IT professionals with a variety of
skills to build and maintain their infrastructure to support and use AI
technologies.

 Civil Military Fusion There is a need to synergise efforts towards


understanding, finding, and implementing AI-based technology by major
players in defence, academia, and civil industry. There is also a
requirement for laying down an adaptive roadmap with priorities to ensure
the best future technologies are incorporated into the armed forces.

 Timelines The above-mentioned capabilities will take time to


develop as they require changes in policy and decision-making. Thus,
there is a requirement to prioritise a clear roadmap defining the goals
required to be achieved in the future. Various goals to be achieved may be
divided into short-term, medium-term, and long-term solutions as given
under Figure 3.

 Short Term (2022-2025)

 Awareness at the technical level.

 Reduction of payload carrying capacity.

 Introduce soldier borne sensors.

 Autonomous robots for route clearance and


Improvised Explosive Device clearance.

 Start maturing policy & improve data storage.

20
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

Figure 3: Timelines and Goals for AI in Indian Armed Forces.

 Mid Term (2025-2030)

 Improvements in technology induction.

 Induction of AI and new-age technology/applications


for semi-autonomous robots.

 Collaboration with commercial sector.

 Long Term (2030 and beyond) The research facilities,


data and the lessons learnt would have matured, thus trials on the
AI-based systems could be carried out extensively and AI
technologies progressed further.

Conclusion

The Indian Army is converging towards IBG-isation as one of the


responses to the progressive changing dynamics of war. The creation of these
forces and tailor-made structures is a right step towards achieving the military and
operational objectives of the nation. As we progress towards learning, weighing
and comparing the pros and cons of employment of AI-based systems in the
Indian Armed Forces, the AI market for India in the civil domain is rapidly rising. As
an example, India was ranked third in the G20 countries for AI-based start-ups.
21
PINNACLE 2022
Strategy to Evolve Artificial Intelligence in Tactical Operations for the Indian Army

The Indian defence forces should realise the dual nature of AI and make
necessary amendments in the policies to enhance the participation of AI-based
start-ups in defence.

Bibliography

1. Artificial Intelligence: Autonomous Technology (AT), Lethal Autonomous Weapons Systems


(LAWS) and Peace Time threats By Regina Surber, Scientific Advisor, ICT4Peace Foundation and
the Zurich Hub for Ethics and Technology Pp 5.

2. Hoadley D,Lucas N, Artificial Intelligence and National Security, April 2018.

3. https:// www. c4isrnet.com/ ome/ 2017/ 09/ 07/army – logistics – integrating – new – ai -
cloud-capabilities/.

4. https://www.drdo.gov.in/sites/default/files/drdo-news-documents/NPC_10-12_April_2021_ 0.
pdf.

5. H Michael, Artificial Intelligence, International Competition, and the Balance of Power,


Texas National Security Review: Volume 1, Issue 3 (May 2018).

6. T Chand, AI and its applications for defence and security forces, Synodos Paper Vol - XII
NO-7/Jun 2018.

7. The Global Race for Artificial Intelligence: Weighing Benefits and Risks, IDSA Issue Brief,
23 Feb 2018 by Munish Sharma.

8. Discussion Paper NITI Aayog National Strategy for Artificial Intelligence June 2018.

9. Report of The Artificial Intelligence Task Force, Ministry of Commerce and Industry (Govt of
India).

10. https:// www. c4isrnet. com/ ome/ 2017/ 09/ 07/army – logistics – integrating – new – ai -
cloud-capabilities/.

11. T Chand,AI and its applications for defence and security forces, Synodos Paper Vol - XII
NO-7 / JUN 2018, PP 6.

12. The Global Race for Artificial Intelligence: Weighing Benefits and Risks, IDSA Issue Brief,
by Munish Sharma.

13. India and the challenge of autonomous weapons R. Shashank reddy.

14. Report of The Artificial Intelligence Task Force, Ministry of Commerce and Industry (Govt of
India) Pp 25.

15. https://searchenterpriseai.techtarget.com/feature/Designing-and-building-artificialintelligence
infra-structure.

22
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

ARTIFICIAL INTELLIGENCE: IMPACT ON WARFIGHTING


AND PERSPECTIVE PLANNING
- Colonel Harish Totade

The nation that leads in AI will be „the ruler of the world


-Vladimir Putin

Introduction

The term "Artificial Intelligence" (AI) generally means that those tasks
which humans could do with organic intelligence will be executed by machines,
viz., identifying a pattern, learning by repetition, making a conclusion, or acting in a
particular manner.1 It is about the study of computations that makes it possible to
perceive, reason, and act. In addition to hardening the mechanisms to counter
threats to important infrastructure like medical assets, power grids, stock markets
etc., AI can optimise our combined abilities to identify, respond, and forecast
threats from a variety of sources. If we turn to AI, the prevailing definitions typically
denote a wide range of capabilities covered by human intelligence that includes
these elements. In a military context, the United States Defence Science Board
states that AI is "the capability of computer systems to perform tasks that normally
require human intelligence." Together, these provide relevant building blocks of a
definition of AI.2

The rapid development of technology and its spread is altering the


character of war and international polity. AI has already permeated a wide range of
fields, both civil and military, making inroads and paving the way for
transformation. However, AI in isolation is not the key element of change; it’s
application in the realm of military routine tasks can hasten decisions, regulate
speed of operations and someday, may push conflict beyond human cognitive
ability.

Specific to the military requirements, the vast applications of Al assure


better early warning models for the onset of conflict, enabling decision makers to
pre-empt and/or prevent it. It can help save the lives of soldiers through better on-
site medical diagnosis. In fact, Al may put fewer soldiers in direct combat by
employing unmanned systems. Al applications may automatically repair faults in
software-driven weapons in real time, which would otherwise take humans a
prolonged period.3 With maturing technology, AI is already shaping new doctrines
and will continue to influence strategy for the future. Conventional weapon
systems powered by AI may destabilise strategic assets and the balance of power
by making the fog of war denser there by creating confusion or escalation,
prompting the start of nuclear confrontation.

23
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

Current Capabilities and Trajectory in Future

AI is already a military reality and since it is not one particular application


but a whole system, as such, it can be said that the armed forces of the world are
seeing it’s inroads in some form or the other. For example, weapon-guidance
systems can make independent decisions irrespective of human input (barring the
setting up parameters before launch) and large datasets can be analysed to
identify patterns. Peculiarly, AI systems with networked computers will have
autonomous decision-making in a rapid, sequential, and uncertain environment.
Intelligent systems adept at drawing inferences on the basis of random data will
likely be on the battlefield and learn from previous actions and from those of
parallel machines.4

At the cutting edge of AI research, there is rapid development in algorithms


for self-learning even with limited data coping with vague and nonlinear data. In
general, ongoing AI research is focused on areas other than defense, such as
civilian vehicle manufacturing, the internet, and healthcare.5 Overall, the
developed capability to categorise asymmetric information and its utility as a
decision-making basis have numerous tactical applications. Early studies promise
that AI will be employed across the spectrum from tactics to strategy.6

It is expected that autonomous systems will be more flexible and deploy


optimally than those guided by human decision-making in the near future. Already,
AI systems can outshine pilots in simulated air-to-air combat.7 The scope of AI will
additionally be expanded to include administration, human resources (postings
and profiling), weapon development, etc. These apparent tactical actions have the
potential to collectively transform a nation’s strategy. AI presently is narrow in the
sense that it can continue in its own field and not expand into other functions on its
own.

Inspite of the recent developments, certain thinkers point out to the


recurrent cycle of optimistic progress in AI, followed invariably by failure to deliver
and a virtual plateau as concepts fail to achieve results in autonomous decision-
making8. However, the full extent of its influence on reshaping armed forces
cannot be predicted with certainty.

Technology and Impact on Strategy

Even though there is considerable development in the field and advent of


many promising technologies are envisaged in the near future, there seems to be
no clear approach to it in most nations as to where and how it will be implemented
in the military. Whether it will be limited to only tactical level applications or be
used in a way that doctrines and strategy will have to be re-written is a matter of
speculation. The military advantage in AI will go to the nation with the most
focused investments and the most comprehensive enabling environment.

24
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

Throughout the history of warfare, it has been seen that the advent of new
technology has a profound impact on strategy. Character of war was transformed
by inventions like gun powder, rifles, tanks, or steam power, but they affected
strategy moderately at best. These changes, however, did not affect the primacy
of fundamental principles like surprise, concentration of force in space and time,
etc.9 Inspite of the previous technological advancements transforming war and
society, they did not change psychology as such, i.e., the primacy of evolved
human mind in strategy.10

Theoretically, the AI system should differentiate between combatants and


civilians based on pre-fed parameters and further choosing to apply lethal force
accordingly. However, these systems, like humans, will always have the potential
for confusion and error. Due to its speed, precision, accuracy, and unbiased
nature, AI is likely to be more favourable for offensive actions. Though the attribute
is equally applicable for defence, the balance is shifted by two factors. By
definition, offensive acts have the initiative, and enemy infrastructure may be
overwhelmed by a mature AI ecosystem. As per Clausewitz, the culminating point
in warfare is a psychological factor in defence as well as attack. Humans are risk
averse, likely to gamble more when losing and will value in-hand possessions
rather than perceived victory.11

Transformation of Key Processes

AI systems will optimise core activities such as reconnaissance,


manoeuvre, and deceive enemy forces through precise fires and rapid
concentration at both tactical and operational levels, increasing lethality and
lowering risk on the battlefield. The developed armies will see their military power
optimised and enhanced exponentially compared to the rest, way beyond the
advancements gained through the information revolution, which will surely change
the power balance.12 They shall have multiple impacts on processes as listed
below:-

 AI will reduce the risks, especially because own casualties will be


lower, and thus lower the threshold for countries that would otherwise
avoid conflict. The risk-averse state may now find it affordable to wage war
due to perceived affordability. Contrarily, AI may negate aggression by a
rogue nation seeking easy gains that are no longer below the interventional
threshold.13

 AI shall shape military actions holistically instead of distinguishing


between conventional and grey zone warfare. The thoughts and methods
for distinguishing the two and exercising restraint need greater deliberation
and definition right from the inception stage.

 Accuracy in decision-making and prognosis of battle scenarios can


be corroborated by AI-enabled war-gaming and advanced simulations.
25
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

Additionally, deterrence capability can be furthered by autonomous


platforms and hypersonic smart missiles.14

 When exploited, AI can filter clutter from large unorganised datasets


and derive actionable intelligence, which may prove mission critical to the
commanders on the ground.15

 The enemy’s Air Defence may be overwhelmed by smart drones,


thereby increasing the effectiveness of own air interdiction operations,
amphibious ground assaults, long-range vectors, and maritime operations
in degraded, contested, and denied environments.16

 The employment of AI in military is not devoid of its challenges. To


be reliable, the systems need to be trained on large datasets.17 It may
predictably overrate the existing threat, transforming the nature and
contours of the situation.18 Such vulnerabilities would derive mostly from
the risk posed by the enemy employing means to deceive, disrupt, or
impair Command and Control (C2) systems19 and hence bring out the need
for inbuilt security.

 Based on the various scenarios and likely future military application,


a brief summary was put up by Niklas Mashur which analyses the
advantages and likely disadvantages of AI.20
Factors Benefits and Potential Advantages Disadvantages and Risks
Strategic  More precise, faster situation  Low crisis stability due to
Decision assessments and analysis. acceleration of decisions.
making  Prejudices can be inherent in
 Offsetting emotions and algorithms.
prejudices.  Problems regarding the balance
of power within states, for example
 Rational behaviour in crisis between the military and the
situations. civilian leadership.
Training and  Personalized training, fair  Overestimation of AI-generated
Organization assessments and promotions. results.
of Armed  More realistic exercises,  Cultural and personnel
Forces manoeuvres and simulations. problems due to incompatibility
 Credible simulations of future between military culture and values
technologies and their applications. held by specialized personnel.
 Military cast system due to
higher technical specialization

Military  More efficient processing of data  Potential dependencies that


Operations from different sources. cannot be replaced in the field.
 Reduction of administrative and  Risks in supply chains due to
staff work through forward-looking lack of inventories and reserves.
logistics.  Unclear whether autonomous
 Reduced risks for troops through vehicles can be used in complex
autonomous logistics. scenarios.
 Improvement of support and  Reduction of strategic stability.
reconnaissance systems.

Advantages and Disadvantages of AI in the Military Field

In other words, AI provides the potential to employ systems lethally,


unconventionally, and asymmetrically without responsibility, given the absence of
26
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

firm international law and ambiguous rules of engagement. Robotic systems when
deployed with Electro Magnetic (EM) jammers can cloud jammers, confuse enemy
Command, Control, and Communication (C3) systems, penetrate hostile systems,
as well as provide layered defence. The doctrines and concepts are likely to
undergo changes consequent to the speed and nature of war due to the presence
of weapons like autonomous robots.21

Roadmap and Implementation

There is a definite requirement in the Indian context to first accept the


requirement of AI in Armed Forces. As per the NITI Aayog Discussion Paper 2018,
the applications of AI in India have referenced utilisation in various sectors but
makes no mention of military applications. However, the current confrontation with
China raises concern and necessity to match research and development (R&D) for
use of AI in the Armed Forces. Simple changes in the present organisation and
planning may be needed, which shall entail:-

 A broad thought on the possible usage and formulation of a tri


services doctrine. The fact that AI is needed in the armed forces needs
acceptance at the highest levels including policy makers.

 Leadership at the strategic level, which appreciates the potential of


AI needs to create policy, allocate resources and ensure charting of a
roadmap. Ultimately, AI technology needs to be involved into products and
processes to solve challenges in the battlefield. This shall also entail
training and adaptations in Tactics, Training and Practises.

 AI advancements will initially occur in bits and processes will


eventually merge to form a system. Hence, the need to have in place
scaling at all levels and intra organisational transparency will be a
prerequisite. This will require creation of a common platform for shared
data, reusable tools, frameworks, software standards, cloud and edge
services.22

 Developing AI into products and process would require employing a


pool of best talent available in the country, not only to harness the existing
systems but also to develop them with the contemporary changes in the
field as well as the changing strategy and operational construct. This will
necessitate the involvement of leading organisations such as the Centre for
Artificial Intelligence and Robotics (CAIR), National Association of Software
and Service Companies (NASSCOM), Army Design Bureau (ADB),
Simulator Development Division (SDD), and others, as well as internal
oversight for ethical application, particularly in weapon-related
infrastructure.

27
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

 Establishment of a focal point to handle all AI-related issues, as well


as a link between the doctrinal approach, development, current strategy,
and most importantly, implementation and integration at all levels23, with a
feedback mechanism for continuous updating and refinement.

 Priority in fielding AI systems, firstly those systems that enhance the


capabilities of humans by replacing strenuous cognitive or physical tasks in
the Indian Armed Forces context and introducing new ways of working.
This requires multi-disciplinary technological development in an iterative
way closely associated with all users.24

 Forming partnerships in both academic and innovative fields is


another key aspect. The leading think tanks in AI and strategic study
organisations need parallel studies to entice new talents in both fields to
match the technology to tactics and vice versa.

 Development in present war-gaming modules and test them in real


situations in parallel to assess their potential for offering advice/planning
susequent or immediate actions/relocating resources/ change in logistics/
predicting enemy actions etc. Development of virtual reality applications
that would reduce the cost of peacetime training while increasing its
realism and facilitating integration.25

 Bringing together all present cyber organisations to focus on the


applications of AI in a phased manner with definite timelines of
development. Concurrently, the technological threshold of rank and file
needs upgradation with careful planning.

 Technology, being fast paced, will need constant upgrades in


hardware as well as software. This has to be addressed in the procurement
model and paced accordingly. The hurdles due to the prolonged loop at
grassroot level have to be mitigated.

Ethical Concerns and Security

The quantum of research in AI technology is nearly similar in volume to that


of its ethical concerns and has centred broadly on the issues of control and
accountability.26 This compounds the problems of system management, technical
expertise integration, and requirement of black or white language demarcating
ethics of AI. In this context, AI systems will inevitably reduce boots on the ground
(though extremely unlikely in the Indian context given the border issues with both
neighbours) and change channels of command. New relationships between a
society in general and its army will likely emerge.27 In any event, it is commonly
agreed, ethically and politically that man has to be in the decision-making loop and
be the executive authority in any event. Robert Work, the US Deputy Secretary of
Defence stated in 2016 that the US Department of Defence (DoD) ‘will not
28
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

delegate lethal authority to a machine to make a decision’ in the use of force.28


The most often repeated question which remains unanswered is that of
accountability in case of a mistake by the machine leading to fatal errors.

Though it is certain that the majority of the algorithms and software on


which AI runs are dual use, with both military and civil contexts, with codes and
machines potentially replicated and commercially used in all fields. Therefore,
once a specific Al application is working, it will be comparatively easy and cheap to
produce similar systems as opposed to the cost of traditional military hardware,
which is inherently expensive.29 There are significant implications for the question
of how production, proliferation, and the use of a wide variety of Al applications
can be controlled by states. Hence, Al is multi-purpose, multifaceted, not
necessarily easily controllable and is likely to have both long-term and short-term
effects that are respectively disruptive and incremental in nature.30

The inclusion of Al in existing military capability can upset the military


balance of power by making traditional systems and doctrines obsolete. The
political costs of going to war are likely to be lower if fewer humans are sent to the
front, thereby reducing an important constraint to going to war. However, immature
Al applications rushed into military platforms can lead to spiral dynamics if they
perform wrong actions due to battlefield environment, similar to problems in
financial trading.31 The study conducted by Research and Development (RAND)
Corporation regarding risks and concerns related to AI in military lists certain
factors repetitively and they are illustrated below in Figure 1.

Figure 1

Conclusion

Hitherto, the introduction of AI into military systems is limited. Yet,


irrespective of the timescale, an adaptive AI for decision-making matters does
29
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

presage a prominent shift in strategy that may be more radical in nature as


compared to the nuclear revolution.32 As compared to humans, the AI scores
majorly in being non-susceptible due to its great speed and accuracy in processing
information. Concurrently, the AI will continuously enhance all tasks by being
devoid of human cognitive tendencies.33

History bears testimony that any technological advantage means little


unless war fighters complement the technology with a comprehensive enabling
environment. The British first invented the tank and led the development of the
technology, however, the Germans developed training, doctrine, concept of
operations and organisational structure to use tanks effectively, resulting in
German tank battalions regularly outmanoeuvring allied forces in the early years of
World War II. As AI transforms military warfare, the advantage will go to the nation
with the most strategically focused investments and most inclusive and
comprehensive environment.

By accelerating decision-making, shortening the Observe, Orient, Decide


and Act (OODA) loop and optimising Command, Control, Communication,
Computer, Information, Surveillance, and Reconnaissance (C4ISR) capability, AI
can provide a cutting edge to the armed forces. However, until solving the mystery
of unpredictability, inflexibility, and brittleness in nature of AI, this phenomenon will
continue to "outpace strategy and human error’.34 Transparency in these decisions
is extremely useful as a confidence and security building measure.35 Until there is
a doctrinal and focused thought with specific time-bound goals that are neatly
implemented, the gap between what could have been possible and what is
available will continue to grow.

Bibliography
1. Summary of the 2018 Department of Defense Artificial Intelligence Strategy, Harnessing AI
to Advance Our Security and Prosperity.

2. Artificial Intelligence and Its Future Impact on Security. Testimony prepared by Dr. Tim
Sweijs for The Committee on Foreign Affairs and the subcommittee on Security and Defense of the
European Parliament. Bruxelles, ro October 2018.

3. Ibid.

4. Sergey Levine, Timothy Lillicrap and Mrinal Kalakrishnan, „How Robots Can Acquire New
Skills from Their Shared Experience‟, Google Research Blog, 10 March 2016,
ttps://research.googleblog.com/2016/10/how-robotscan-acquire-new-skills-from.html.

5. Kenneth Payne. Op cit pp 14.

6. Kareem Ayoub and Kenneth Payne, „Strategy in the Age of Artificial Intelligence‟, Journal of
Strategic Studies, vol. 39, nos 5–6, 2016, pp. 793–819, https://doi.org/10.1080/01402390.2015.
1088838. See also Michael Horowitz, „Artificial Intelligence, International Competition and the
Balance of Power‟, Texas National Security Review, 15 May 2018, https://tnsr.org/2018/05/artificial-
intelligenceinternational-competition-and-thebalance-of-power/; and Paul Scharre, Army of None:
Autonomous Weapons and the Future of War (New York: W.W. Norton, 2018).

30
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

7. Nicholas Ernest et al., „Genetic Fuzzy Based Artificial Intelligence for Unmanned Combat
Aerial Vehicle Control in Simulated Air Combat Missions‟, Journal of Defense Management, vol. 6,
no. 1, 2016, https://doi.org/10.4172/2167-0374.1000144.

8. Payne loc cit.

9. See Beatrice Heuser, The Evolution of Strategy: Thinking War from Antiquity to the Present
(Cambridge: Cambridge University Press, 2010).

10. Payne op cit pp 4.

11. Kenneth Payne, „Prospect Theory and the Defence in Clausewitz‟s “On War”‟, paper
presented to the annual meeting of the International Studies Association, Atlanta, GA, 16March
2016, http://web.isanet.org/ Web/Conferences/Atlanta%202016/ Archive/968b741d-130c-4912-a4ee-
997345a57ce1.pdf.

12. Ibid.

13. Ibid.

14. Linda Lastovych. Op cit pp 5.

15. James s. Johnson. Op cit pp.

16. Linda Lastovych op cit pp 6.

17. Michael C Horowitz and Paul Scharre, „AI in international Stability: Risk and Confidence
Building Measure (Centre for New American Security 2021) pp5.

18. James s. Johnson, „Artificial Intelligence and Future Warfare: Implications for international
Security (2019) pp 35 Defence & Security analysis pp147.

19. Mark Fitzpatrick, „Artificial Intelligence and Nuclear Command & Control 2019 pp 61.

20. Niklas Mashur. AI in military Enabling Applications. CSS Analyses in Security police no 251
Oct 2019 pp 3.

21. Ibid.

22. Summary of the 2018 Department of Defense Artificial Intelligence Strategy op cit pp.

23. The US DoD has already established a Joint Artificial Intelligence Center (JAIC) to
accelerate the delivery of AI-enabled capabilities, scale the Department-wide impact of AI, and
synchronize DoD AI activities to expand Joint Force advantages.

24. Ibid.

25. Atul Pant, Artificial Intelligence The Visible Path, IDSA Occasional paper. No 49, Aug 2018.

26. Ugo Pagallo, „Robots of Just War:A Legal Perspective‟, Philosophy &Technology, vol. 24,
no. 3, September 2011, pp. 307–23, https://doi.

27. Ibid.

28. Linda Lastovych op cit pp 6.

31
PINNACLE 2022
Artificial Intelligence: Impact on Warfighting and Perspective Planning

29. Dr. Tim Sweijs. Op cit pp 5.

30. Ibid.

31. Ibid.

32. Ibid.

33. Payne op cit pp 22.

34. James s. Johnson. Op cit pp.

35. Draft Guidelines on Developing National Defense Policy And Doctrine Papers ("White
Papers") (Approved by the Committee on Hemispheric Security at its meeting held on October 22,
2002).

32
PINNACLE 2022
ARTIFICIAL INTELLIGENCE IN CYBER SPACE:
TOWARDS MODERN WARFARE
- Miss Khyati Singh

Introduction

The AI currently available to us is called „Narrow AI‟, and its evolution


would lead us to „General AI‟, which has the power to revolutionize military
operations. However, the presently available AI is more evolutionary than
revolutionary which facilitates decision making. Currently, AI is being developed,
and used in cyberspace & military operations in all three domains - air, water and
land, and for strategic placement of weapons which also involves decision-making.
Despite the multifaceted use, the strategic planning in and around AI has been
obscure and lacks quality R&D. A repetitive mistake has been labeling slight
technological modifications as AI. This has created an environment of
misinformation where basic coding and system advancement is dubbed as “AI”.
Consequently, it broadens the bracket of AI without any depth, and keeps the
human resource involved in petty technological projects. The core aspect of AI
being a „replica of human mind‟ has been relegated to the sidelines. The sole fact
that makes AI „revolutionary‟ is that it aims to „replicate human brain‟. The
cognitive skills are what puts it apart from every technological ace the world has
developed till date. As scholars worry about and populate the existing literature on
a possible „AI arms race‟, they miss out on the basic questions of AI, and its
implication. Narrow AI, is comparatively less complex, and has fair share of
technological limitations along with the constraint to undertake multitasking.
However, General AI would fix these challenges as it will be organized along more
complex structures that will be of great advantage to military operations and
intelligence. For instance, Harpy drone is being used by Israel to counter enemy
operations, Governments are utilizing AI to extract big data for counterterrorism,
while the New Zealand Navy is using it for logistics. China too has worked on
drone swarm technology that is beneficial in bypassing the enemy defense base.1

Weaponization of AI in Cyberspace

The speculation of AI being weaponized and causing drastic changes has


become a reality of late. Weaponization altered conventional warfare weapons
used in the air, sea, land, and space. Moreover, this weaponization of nuclear and
chemical materials has been linked to space usage and climate manipulation. The
presence of weaponized AI in cyberspace is alarming. It corrupts the system and
adds malicious data that can be used to hinder its smooth functioning and send
across rigged messages. Weaponzied AI has been defined as ―Malicious AI
algorithms that can degrade the performance and disrupt the normal functions of
benign AI algorithms, while providing technological edge attack scenarios in both
cyberspace and physical spaces.‖2

PINNACLE 2022 33
Artificial Intelligence in Cyber Space: Towards Modern Warfare

Cyberspace operates on data, and with its exponential growth, the data it
works upon will also increase. This will make it next to impossible to keep it in
check through a human agency alone. This is where the element of AI comes in.
AI helps in assuring the security of cyberspace along with managing the enormous
data. However, tampering with the hardware and software can be extremely
dangerous. For instance, autonomous systems based on AI models can be used
to disguise or divert people. Researchers at UC Berkeley firm developed a model
based on this feedback. They devised a false stop sign for the autonomous driver,
which confused fellow drivers and almost caused a road accident. Inducting this
prototype into the military can be worrisome. Enemies can hack into the systems
of army vehicles and alter the settings.

Weaponized AI cyberspace can either be directly integrated into battles or


be inducted into military operations that deal with computing, data analysis,
systems operation, decision making, etc. The latter aspect is more prevalent given
the current scenario since it involves real-time decision making, maintenance of
prominent supply chains and doing jobs that humans refrain from doing. All these
factors together affect the power dynamics of conventional warfare. The power of
"big data analysis" is crucial to any operation. This big data is referred to as
„mass‟, and the bigger the 'mass', the better performance AI has. Another aspect
that is often considered easy with AI is decision making. However, the present-day
AI does not afford that comfort. Human agency is integral to the process. The aim
of AI remains to replicate human brains with the same cognitive capabilities, but
that is a far-flung goal. Present day reality requires human intervention because
machines do not understand the human codes of morals and rationality. Their
algorithms work on a designated pattern. Hence, when it comes to cyberspace,
human agency has to be given more attention and the required edge.

Conventional vs Modern Cyberattacks

Conventional cyber-attacks are devoid of AI, whereas the modern cyber-


attacks are backed with AI. The conventional methods, or the most common ones
are Distributed Denial of Service attacks (DDoS), Denial of Service attacks (DoS),
Phishing, Man in the Middle (MitM) attacks, injections of Structured Query
Language (SQLI) and Cross-site Scripting (XXS), Malware attacks, Ransomware
attacks, Eavesdropping attacks, Scarce attacks, etc.3 A variety of these are also
being used in the Ukraine-Russia war. However, they did not qualify as AI
powered attack. AI powered attacks include data misclassification, data analysis
and synthetic data generation. The technology deployed to tweak these factors
can be applied to various other warfares.

Trail of events for AI development

AI advancements are not new and have been in the works since the 1950s,
However, the current rate of progress is unprecedented. The Defence Advanced
Research Projects Agency (DARPA) 4 led the research in AI and paved the way for
34
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

nuclear powered AI supersonic jets. Likewise, the induction of AI into cyberspace


is in the pipeline and just one step away from the breakthrough that everyone
looks forward to. It was with the development of AI in the military with projects like
the "Survival Adaptive Planning Experiment" that was tested for AI decision
making. It was concluded that though AI decision making will become feasible with
time, the real challenge will arise when the power of decision-making rests both
with humans and AI5. It might create a conflict that either collides or diffuses the
actions. Moreover, the lack of an accepted definition for addressing what
"autonomous" means has been a cause of concern. Initially, the autonomous tag
meant little human intervention, but gradually, as the advancements in AI are
taking place, it is being envisioned as something that has no human variable to it.

In addition, the other issue that encircles the development of AI is that of


ownership and creation. It was the private sector that led the development of AI
and not the government. Private firms have a model that is at ease compared to
the government, which falls prey to regulations and turf battles, along with the
money that goes into these developments. Private agencies could produce the
same or better results in little time as compared to government agencies. This
regulation-free operation makes it easy for the private sector to produce and sell
the technology, which might end up with either a rogue state or a terrorist
organization.

Even as the trend towards exploitation of AI for military purposes cannot be


rendered, there is a need to manage it effectively. It has already penetrated
various domains of operations, and it now focuses on cyberspace. Most AI-
powered attacks in cyberspace take place by inducing noise in machine learning
algorithms that pave the way for adversarial machine learning. It corrupts the
network and creates a ruckus. This can be defended by modifying the neural
network that has additional classifiers along with the data that constitutes the
defensive domain. 6

Mr Vasisht Duddu, a researcher from the Indraprastha Institute of


technology7 undertook a survey in 2018 that studied adversarial machine learning
in cyber warfare. He identified a series of threat models and defensive strategies.
The adversarial machine learning attacks included model inversion attacks,
equation solving attacks, path finding attacks, evasion attacks, member inference
attacks, transferability attacks, and black box attacks. Furthermore, he pinpointed
the vulnerabilities in machine learning algorithms, encompassing supervised,
unsupervised, and reinforcement learning algorithms. The defence mechanisms
available for these attacks were privacy-preserving algorithms and processing
techniques that safeguard machine learning-based algorithms.

Mr Naveed8 also conducted a survey in September 2019 that charts the


countries that are leading the AI battle in cyberspace. China topped the list
followed by Iran, Singapore, the US and India. However, majority of the research
that these countries and their universities were doing aimed at data mining 9,
35
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

network traffic classification10, fraud detection11, adversarial machine learning12,


and anomaly detection13. The following figure14 shows what has been dubbed “AI
hotspots”. These are the countries that are leading the research on AI in
cyberspace.

Researchers came across Generative Adversarial Network (GAN)15 that


proved to be a breakthrough in deep learning. GAN along with its medical imaging
applications, works like a neural network model where two networks are trained at
the same point of time. One of it focuses on image generation while the other on
discrimination. For AI powered cyber- attacks, GAN is of immense importance.
The main function of GAN falls under data generation in terms of visual, textual,
and audio content. The range of applications include audio synthesis, super
resolution, text mining, drug discovery, video in painting, synthesizing data for
training which can be used for further training deep networks.

GAN has two prominent agents, namely, generator and discriminator.


Both of them are deep networks but have different loss functions. The generator
network first learns the dissemination of data and then produces samples for the
discriminator network. Further, the discriminator is given two kinds of samples: one
that comes from the generator and the other that is the original data. Therefore, it
is the job of the discriminator network to identify whether the generator sample has
been sourced from the original network or from the generator network. The training
is devised to train the generator to trick the discriminator by replicating original
samples, while the discriminator is trained to identify the original sample. This is an
end-to-end training network that ends when it acquires Nash equilibrium16. In the
context of networks, this equilibrium works when the loss of the generator, that is,
the number of times it failed to trick the discriminator, is equal to the loss function
of the discriminator, that is, lack of discrimination between the two samples.
36
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

However, reaching such an equilibrium is a tall order since the loss functions
oscillate. The following picture17 is a pictorial representation of the GAN diagram.

AI Induced Cyberattacks

Research has tried to identify machine learning strategies' relationships


with cybersecurity strategy and functioning. Machine learning relies on algorithms
that work on the feedback gained from previous experiences to comprehend
repeating patterns in place of programming the patterns themselves. Machine
learning helps find solutions to the same problems. 18. Two types of algorithms that
can be deployed in machine learning are supervised and non-supervised
algorithms.

To solve the issue of cybersecurity, military bases can deploy these


methods of machine learning. Militaries can induct regression, classifications,
generative models, clustering, and association rule learning. By applying
regression in cybersecurity defensive measures, fraud or detection of any
suspicious activity can be outlined. Classification helps in organising or
categorising data based on the training that is being provided that contains a
category for known membership observations. Whereas clustering is a tool of
unsupervised learning using which data points can be employed or populations
can be grouped. Meanwhile, generative models stimulate real data by devising a
model that creates a list of information parameters that evaluate specific
applications for injected vulnerabilities.

Machine learning also plays it‟s part both as an offensive and defensive
actor. On the offensive end, cyber security tools like spearfishing, evasive malware
and botnets can be used. While on the defensive front, machine learning tools like
malware detection is being used. The hacker needs to crack the code just one
time, and that one single access can help in corrupting the entire system. Military
intelligence needs to be on their toes 24x7 and must work with 100% accuracy to

37
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

block these attacks. To facilitate this, machine learning can imitate human
behavior and prepare a database that gives out possible services based on the set
of behavior. This can even help in preventing attacks that have not yet taken
place. However, leakage of data is very much a possibility in machine learning
systems. Big companies like Google and Amazon have had such experiences. A
shadow training technique has been devised to resolve this issue. This helps in
filtering the actual data from nosy interference data.

AI powered tools which deploy data analysis for offensive operations in


cyberspace are numerous. The most common ones are DeepHack19- a tool that
creates injection attack patterns that can be applied to database, DeepLocker 20-
emulates an Advanced Persistent Threat (APT) to facilitate the launch of complex
cyber-attacks, GyoiThon21- which is used for gathering information and automatic
exploitation, EagleEye22- uses facial recognition algorithms for information
reconnaissance for social media, uriDeep23- generates domain names that are
fake but can be used in various attack scenarios and works as a proxy, Deep
Exploit24- automates Metasploit for gathering information, exploitation and post
exploitation, and Deep Generator- which works for web applications by producing
injection attack patterns. These tools that are completely AI driven can help
counter the hackers that feed false information into the machine learning models,
or at times evade detection using technology codes that crack through the
systems. To keep pace with such nefarious activities, deployment of AI tools in
military cybersecurity space is extremely crucial.

On the audio front, deep fakes are increasingly becoming an attack force to
reckon with. An attacker no longer needs a voice actor or performer to mimic or
copy the desirable voice. Deep Fake generates the exact voice that can be used
to give commands. In warfare, or in military operations in general, orders come via
various audio devices owing to the different locations of operations. Under this
scenario, fake voices imitating the actual person in charge can cause confusion
throughout the entire operation. These kind of fake audio-induced attacks are
known as „vishing‟25.

Future Course of Actions

In cyberspace, it is an open war of AI vs AI. It would either end in a draw or


might end up favouring the one with greater might. Offensive AI in cyberspace
looks for the loopholes and penetrates vulnerabilities to cause harm. At the
defensive level, the proposed AI would figure out these loopholes to cover up
vulnerabilities and keep them from further damage. This, in the sense, is also a
question of timely interjection. For instance, the Defence Advanced Research
Projects Agency (DARPA) cyber grand challenges26 make AI-based cyber
defenders and attackers compete. This helps the government track the possible
threats and their feasible solutions. The Indian military can too adopt such
innovative methods that help in strengthening their systems. Likewise, in
adversarial machine learning, GAN can be used both in defense and offense. The
38
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

biggest challenge that AI faces is the acquisition of required data that facilitates
the creation of accurate models. This the military can resolve by expanding their
resource base and incorporating more data analysis tools. Furthermore, with the
introduction of new and advanced technologies like 5G/6G, better infrastructure
will be required.27

Another set of security systems that can help counter security breaches in
cyberspace are: signature-based, which try to map out all the cyber-attacks with
those attacks' specific signature or pattern; anomaly-based, which keep track of
regular machine behaviour and raises an alarm if there is any deviation from the
normal sense; hybrid based, is a combination of the previous two, that is,
anomaly and signature. It keeps an account of the intrusion rate and thereafter
tries to bring down the range of unknown attacks.28

Cyber attackers often exploit the overfitting attribute of GAN where the
algorithms learn the distribution of the training data accurately; however on the
validation set they lack precision. This overfitting is easier to deal with in the case
of convolutional neural networks, Boltzmann machines and auto encoders, but
when GANs are involved, it becomes rather challenging. Hackers exploit this
aspect to produce data to foul the networks. The way to deal with such a threat is
through „network regularization‟ which aims at developing sophisticated
regularisation. Moreover, a generalised model can be more helpful in mitigating
these odds.

It is also important to secure AI classification models that are generally


exposed to „Evasion‟, „Poisoning‟ and „Stealing‟. In „Evasion‟, the data that works
as an input for the AI algorithm is manipulated; this allows it to bypass the
classification mechanism. Likewise, in „Poisoning‟, the training data is tampered
with to affect the classification pattern. Lastly, in „Stealing‟, the input and output
analysed by the AI algorithm are accessed to understand the model properties and
then create a model that counters those safeguards. Deep-pawn29 is one such
system that was developed in 2016 and can be deployed to tackle the
aforementioned attacks.

These attacks can be on multiple levels, but the military of any country, for
example, needs to recruit AI measures for cyber defence. These include anti-
phishing measures30, kill chains, attack visualization, etc. Phishing is the most
prevalent method of stealing data using the internet. Military operations rely on
information, and during wartime, any information leakage proves to be detrimental
to the entire operation. This needs to be tackled at all levels. Researchers have
tried to develop machine learning algorithms that train the model to detect phishing
attacks. The current AI-based techniques include the Synthetic Minority Over-
Sampling Technique (SMOTE) algorithm for detecting phishing. Furthermore,
there are attacks that are taking place on the grid systems and essential
infrastructure. These are generally controlled by human operators, but has its
limitations, and are generally erroneous. It is important to devote resources to the
39
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

development of effective technologies to counter such attacks. The attacks on


cyberspace have started penetrating the system, and with AI, the devastation
would be unparalleled.

Conclusion

It was initially just the development of AI that the world was dealing with,
but as technology climbed the ladder of development, it got clubbed with cyber
space as well. Cyber-attacks are not new to humans. What is new about them is
their clubbing with AI. This has unfolded a new dimension, where the attackers are
accessing the capabilities faster than the defenders. Moreover, the fact that AI has
not been autonomous to the extent it was believed to be, has further made it
difficult to define a definite trajectory. Therefore, timely involvement, active
measures, and sincere R&D concerning AI in cyber space is the need of the hour.
Militaries are primarily trained in conventional warfare. It is crucial to understand
that, with the changing dynamics of war, it should be our obvious action to
upgrade and update our modes of operation. To provide more cyber training to the
personnel, develop systems and machine learning algorithms that could safeguard
the system. In a war of AI against AI, only technology will win.

Bibliography

1. China Releases video of new barrage swarm drone launcher. 2020,


https://www.forbes.com/sites/davidhambling/2020/10/14/china-releasesvideo-of-new-barrage-swarm-
drone-launcher/#484d44822ad7.

2. Weaponized AI for cyber attacks Muhammad Mudassar Yamin a,∗ , Mohib Ullah a , Habib
Ullah b , Basel Katt.

3. Yamin Muhammd Mudassar, Katt Basel, Kianpour Mazaher. Cyber weapons storage
mechanisms. In: International conference on security, privacy and anonymity in computation,
communication and storage. Springer; 2019, p. 354–67.

4. Defense advanced research projects agency. 2019, https://www.darpa.mil/.

5. Geist Edward Moore. It’s already too late to stop the AI arms race—We must manage it
instead. Bull At Sci 2016;72(5):318–21.

6. Li Jian-hua. Cyber security meets artificial intelligence: a survey. Front Inf Technol Electron
Eng 2018;19(12):1462–74.

7. Duddu Vasisht. A survey of adversarial machine learning in cyber warfare. Def Sci J
2018;68(4):356–66.

8. Abbas Naveed Naeem, Ahmed Tanveer, Shah Syed Habib Ullah, Omar Muhammad, Park
Han Woo. Investigating the applications of artificial intelligence in cyber security. Scientometrics
2019;1–23.

9. Bekerman Dmitri, Shapira Bracha, Rokach Lior, Bar Ariel. Unknown malware detection
using network traffic classification. In: 2015 IEEE conference on communications and network
security (CNS). IEEE; 2015, p. 134–42.

40
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

10. Ullah Mohib, Ullah Habib, Khan Sultan Daud, Cheikh Faouzi Alaya. Stacked lstm network
for human activity recognition using smartphone data. In: 2019 8th European workshop on visual
information processing (EUVIP). IEEE; 2019, p. 175–80.

11. Abdallah Aisha, Maarof Mohd Aizaini, Zainal Anazida. Fraud detection system: A survey. J
Netw Comput Appl 2016;68:90–113.

12. Biggio Battista, Roli Fabio. Wild patterns: Ten years after the rise of adversarial machine
learning. Pattern Recognit 2018;84:317–31.

13. Ullah Habib, Altamimi Ahmed B, Uzair Muhammad, Ullah Mohib. Anomalous entities
detection and localization in pedestrian flows. Neurocomputing 2018;290:74–86.

14. Abbas Naveed Naeem, Ahmed Tanveer, Shah Syed Habib Ullah, Omar Muhammad, Park
Han Woo. Investigating the applications of artificial intelligence in cyber security. Scientometrics
2019;1–23.

15. Yi Xin, Walia Ekta, Babyn Paul. Generative adversarial network in medical imaging: A
review. Med Image Anal 2019;101552.

16. Nash John F, et al. Equilibrium points in n-person games. Proc Natl Acad Sci
1950;36(1):48–9.

17. https://developers.google.com/machine-learning/gan/gan_structure.

18. Chachra Anjali, Sharma Deepak. Applications of machine learning algorithms for
countermeasures to cyber attacks. 2019, Available at SSRN 3370181.

19. BishopFox/deephack: PoC code from DEF CON 25 presentation. 2020, https:
//github.com/BishopFox/deephack.

20. CyberWarefare/DeepLocker: DeepLocker - Deep learning based malware. 2020, https://


github.com/Cyber Warefare/DeepLocker.

21. Gyoisamurai/GyoiThon: GyoiThon is a growing penetration test tool using Machine


Learning. 2020, https://github.com/gyoisamurai/GyoiThon.

22. ThoughtfulDev/EagleEye: Stalk your friends. Find their Instagram, FB and Twitter Profiles
using Image Recognition and Reverse Image Search. 2020, https://github.com/
ThoughtfulDev/EagleEye.

23. Mindcrypt/uriDeep: Unicode encoding attacks with machine learning. 2020,


https://github.com/mindcrypt/uriDeep.

24. Machine_learning_security/DeepExploit at master. 13o-bbr-bbq/machine_ learning_security.


2020, https://github. com/13o-bbr-bbq/machine_learning_ security/tree/master/DeepExploit.

25. Griffin Slade E, Rackley Casey C. Vishing. In: Proceedings of the 5th annual conference on
information security curriculum development. ACM; 2008, p. 33–5.

26. Avgerinos Thanassis, Brumley David, Davis John, Goulden Ryan, Nighswander Tyler,
Rebert Alex, et al. The mayhem cyber reasoning system. IEEE Secur Privacy 2018;16(2):52–60.

41
PINNACLE 2022
Artificial Intelligence in Cyber Space: Towards Modern Warfare

27. Ali Sher, Ahmad Ayaz, Faheem Yasir, Altaf Muhammad, Ullah Habib. Energyefficient RRH-
association and resource allocation in D2D enabled multi-tier 5G C-RAN. Telecommun Syst 2019;1–
15.

28. Yamin Muhammad Mudassar, Katt Basel, Sattar Kashif, Ahmad Maaz Bin. Implementation
of insider threat detection system using honeypot based sensors and threat analytics. In: Future of
information and communication conference. Springer; 2019, p. 801–29.

29. Deep-pwning - Metasploit for Machine Learning. 2020, https://www. kitploit.com/2016/


11/deep-pwning-metasploit-for-machine.html.

30. Sundararajan Aditya, Khan Tanwir, Aburub Haneen, Sarwat Arif I, Rahman Shahinur. A tri-
modular human-on-the-loop framework for intelligent smart grid cyber-attack visualization. In:
SoutheastCon 2018. IEEE; 2018, p.1–8.

42
PINNACLE 2022
FUTURE OF ARTIFICIAL INTELLIGENCE APPLICATIONS IN
THE INDIAN ARMY COMMUNICATION NETWORK
- Colonel PS Mehta

“Artificial Intelligence is about replacing human decision making with more


sophisticated technologies.’’
- Falguni Desai

Introduction

Artificial Intelligence (AI) was introduced as an academic discipline in 1955


and, since its introduction, has experienced several waves of optimism followed by
periods of disappointment. After ‘Alpha Go’ defeated a professional Go player in
2015, AI once again attracted widespread global attention. AI techniques have
now become an essential part of the industry and are being used to extensively
solve challenging problems in software engineering, communication technology,
and computer science.

AI is being introduced in the defence communication sector in a big way,


with the aim of providing new services, improving network efficiency, and
enhancing user experience. Some pioneering work has already been carried out in
the defence communication field using AI, and its advantages in learning,
understanding, and reasoning have been used to develop Software Defined
Networks (SDNs) and Network Functions Virtualization (NFV). In addition, Deep
Packet Inspection (DPI) and Service-Aware Network technologies are in the
process of development.

Inclusion of AI in Army Communication Network

The Army Communication Network has a varied set of needs that include
diversity, security, and ruggedisation. With the smart soldier concept being
introduced in armed forces across the world, it is all the more necessary to
introduce AI in the army communication network. This will assist the soldier in
taking timely decisions based on inputs received through communication devices
or equipment. The outcome of an operation or battle will depend on the availability
of better and more advanced communication equipment, to equip the soldier with
all the necessary inputs to take a timely decision to influence the outcome of the
battle. In addition, a lot of surveillance and intelligence data is available as inputs
to a commander. It is important to sort the data into useful and generic information
so that only that information is fed to the commander that is relevant to him, so as
to assist him in passing necessary directions timely to his subordinates. Trends in
the field of communication, which necessitate the inclusion of AI in the Army
Communication Network, are given in the succeeding paragraphs.

PINNACLE 2022 43
Future of Artificial Intelligence Applications in the Indian Army Communication Network

 Characterised Requirements With the increasing number of


users and growing size of the Army Communication Network, differences in
preferences, habits, and the informational needs of subscribers are
gradually exposed. The demand for customised networks and services is
now becoming stronger with the varied requirements of data users. In the
near future, there will be complex communication network requirements
whose implementation would be unmanageable without an intelligent tool.

 Multimedia Services Army users have now also become


information producers, as well as information consumers and are
producing more and more information. User generated content increases
traffic at an unbelievable speed. Under these circumstances, both storage
and transmission are great challenges. The inclusion of AI bolsters our
ability to handle this challenge in the Army Communication Networks.

 Precision Management The various dimensions and granularities


in today’s wireless traffic models should be considered in the Army
Communication Networks. With the development of the technologies of
network function virtualization and software-defined networks, the
management of the communication network has become more precise.
Virtualization is not only at the level of network elements but also at the
level of components such as the CPU, memory, ports, bandwidth, etc. AI
based technologies will allow operators to setup on-demand networks.

 Predictable Future The expansion of data requirements and


increasing numbers of end users has ensured that the gap between the
peaks and troughs of Army Data Network usage is becoming greater. It is
therefore required to predict the future status of Army Data Networks more
accurately to satisfy users’ demands and improve their experience.

 Network Performance The increase in network equipment and


user terminals, the expansion of network size, the increase in the number
of users, and the complexity of the Army Communication Network have
resulted in the network management becoming more difficult to maintain
with an acceptable Quality of Service (QoS). In addition, due to expanding
capacity by introducing more equipment as part of Project Network for
Spectrum (NFS) in the Indian Army, there is a requirement to enhance
network performance with smart tools and intelligent technologies enabled
through AI.

 More Attention to Security and Safety Data breach and Cyber


Security violations are growing and becoming severe in the Army
Communication Network. These events have resulted in significant
consequences, including compromise of operational data. AI can be
introduced into several layers to establish strong security protection and

44
PINNACLE 2022
Future of Artificial Intelligence Applications in the Indian Army Communication Network

behavioural analysis based on machine learning which will significantly


improve the ability of network detection attacks, automatic analysis of data,
and the identification of relationships between isolated behaviours.

Advantages of AI in Army Communication Network

 Abilities of Learning Intelligent decisions are required to manage


dynamic traffic to accurately describe the Army Communication Network
traffic characteristics. Fortunately, AI has entered the cognitive age wherein
deep learning can be exploited. Through deep learning, the machine
system can use the existing training data to process large amounts of data
through data mining. AI can be utilised to carry out Army data traffic
management and maintenance. Through these efforts, the accuracy of
analysis can be enhanced and the intelligent management of Army
Communication Networks can be realized.

 Abilities of Understanding and Reasoning Due to the changing


dynamics of the Army Communication Network system, the state of
information source may have changed when it is transmitted to the network
management system. Therefore, the network management can only know
the local state information without knowledge about the system's internal
state. Machine learning happens to have the strength to deal with this kind
of fuzzy logic and uncertainty in reasoning. In order to make the
classification or prediction easier, deep learning constructs a multi-hidden
layer model and uses the hierarchical network structure to transform the
feature representation of the sample into a new feature space, layer by
layer. In particular, AI does not need to describe the mathematical model of
the system accurately and therefore has the ability to deal with the
uncertainty or even "unknowability" that is peculiar to the Army
Communication Network.

 Ability of Collaborating The structural complexity of Army


Communication Network is rapidly increasing as the network expands in
both scale and size. Concepts such as distribution and hierarchy are often
talked about in network management. Management tasks and controls are
distributed to the entire network. As a result, we must deal with issues such
as task distribution, communication, and collaboration between Army
Communication Management Nodes. If we introduce the multi-agent
collaboration of distributed AI into the Army Communication Network
Management, we can expect the ability to collaborate between network
managers distributed at every layer.

 Use of AI in Management, Monitoring and Maintenance of Army


Communications Since the era of Time Division Multiplexing (TDM)
automatic switched to IP based communication on networks, it has been
the pursuit of the Army Communication Engineers to introduce intelligence
45
PINNACLE 2022
Future of Artificial Intelligence Applications in the Indian Army Communication Network

into network operations, management, and maintenance.


Software–Defined Networking (SDN), Network Functions Virtualization
(NFV), etc, coupled with integrated network management systems, can
directly issue orders which can be executed by network equipment, and it
is possible to realise real-time monitoring of Army Communication
Networks and services. To monitor the real-time information of the Army
Communication Network, Deep Packet Inspection (DPI) system can collect
information such as the running state of network equipment, the usage of
resources, and the quality of services. With the big data obtained from the
DPI system, the AI system can rapidly analyse and find if there are or will
be abnormity within the information. For example, if AI system finds a burst
of continuous traffic, it can suspect a Distributed Denial of Service (DDoS)
attack in the network, analyse the package characteristics and orchestrate
an inspector collaboration task to drop all packages with the suspected
characteristics to avoid damage.

 AI in Traffic Prediction AI can be applied to Army


Communications traffic demand prediction during the planning process,
where it is used to predict and analyse traffic demand. Traffic tendencies
can be divided into two types: short-term tendencies, such as temporary
traffic increase during motion exercises; and long-term tendencies, from
which anomalous tendencies such as temporary traffic increases during
events have been removed. In AI-driven traffic prediction, we make AI learn
the short and long term traffic tendencies that have totally different factors
causing traffic fluctuations and mechanisms in order to predict both short
and long term traffic demands.

 AI in Spectrum Management Spectrum is a valuable asset in


Defence Communication Networks and needs to be utilised judiciously.
Piecemeal optimization of applications such as spectrum monitoring are
manpower intensive. It entails efforts to hand-engineer feature extraction
and selection that often take months to design and deploy. By employing
powerful machine learning algorithms, AI can be utilised to efficiently carry
out assigned tasks. In addition, the design complexity of Radio Frequency
(RF) systems can be reduced by improving RF parameters such as
channel bandwidth, antenna sensitivity, and spectrum monitoring.

Conclusion

Indian Army recognises the qualitative edge AI systems will provide in near
future, and will assist commanders faced with unconventional adversaries in high-
speed engagements. AI system will augment analysis and decision-making
capabilities of commanders and speed up learning and improve their ability to act
with discretion, accuracy and care under uncertain and changing conditions.

46
PINNACLE 2022
Future of Artificial Intelligence Applications in the Indian Army Communication Network

Use of AI systems will transform traditional Army Communication Network


into intelligent network. Predictive mechanisms will enable users to take faster and
better data driven decisions. Intelligence built on huge amount of historical network
data will provide a pattern for network anomalies. Using AI models, Army Data
Network administrators can detect network failures, forecast traffic patterns,
understand traffic congestion, and build intelligent security. Use of AI in Army
Communication Network, thus, will enable commanders at all levels to obtain
timely information and further disseminate information to subordinates, based on
analysis available using AI systems to ensure timely action.

Bibliography

1. G. E. Hinton, R. R. Salakhutdinov, Reducing the Dimensionality of Data eith Neural


Networks, Science, Jul. 2006, Vol. 313:504-507

2. N. Kojic, et al. Neural Network for Optimization of Routing in Communication Networks,


Facta Universitatis (NIS) Ser: Elec. Energ. Aug. 2006, Vol. 19, no.2:317-329.

3. SUI Dan, JIN Xian-Hua, Network Cognition Control Method Based on Artificial Intelligence,
Computer Simulation, Sept. 2011, vol. 19, no.2:317-329.

4. Sandra Sendra, et al, including Artificial Intelligence in a Routing Protocol Using Software
th
Defined network, ICC2017: wt04-5 IEEE International Workshop on Smart Communication
Protocols and Algorithms (SCPA 2017).

5. Sahebu, K.M. Artificial intelligence approach to planning and managing communication


networks, International Conference on Electromagnetic Interference on Electromagnetic Interference
& Compatibility, 2002: 193-202.

6. Open Networking Foundation (ONF), Software-defined Networking: The New Norm of


Networks [EB/OL], http://www.opennetworking.org/images/stories/downloads/sdn-resources/white-
papers/wp-sdn-nownorm.pdf.

7. Chiosi M. et al. Network Functions Virtualization: An Introduction, Benefits, Enablers,


Challenges and Call for Action [EB/OL], http://www.etsi.org.

8. M.Al-Hisnawi, M. Ahmadi, Deep Packet Inspection Using Quotient Filter, IEEE


Communications Letters, 2016, 20 (11):2217-2220.

9. XU Guibao, A technological architecture of artificial intelligence, Telecommunication


Network Technology, Dec. 2016, no. 12:1-6.

10. J. Holland, Can there be a unified theory of complex adaptive system?, in: The Mind, the
Brain and the Complex Adaptive Systems, H. Morowitz and J. L. Singer, Eds., Addison-Wesley,
1995.

47
PINNACLE 2022
EMPLOYMENT OF ARTIFICIAL INTELLIGENCE
TO ENHANCE EFFICACY OF ARTILLERY
- Major Sharad Sagar Joshi

Introduction

In wars of the future, the side with shorter and effective Observe, Orient,
Decide and Act (OODA) loop will be able to achieve effect based application of its
lethal and non-lethal assets to break the enemy’s will to fight. A shorter OODA
loop can be achieved by ensuring seamless command and control architecture
with targeting capability of firepower assets providing the cutting edge. The
ultimate role of artillery aims at effective application of all firepower assets to
achieve the desired degree of neutralization, suppression and destruction. This
entails multifarious tasks, primarily including rapid deployment, handling of Artillery
Target Intelligence (ATI), establishment of surveillance grid, target acquisition,
target analysis, allotment of most suitable firepower asset for engagement and
Post Strike Damage Assessment (PSDA). There are well established procedures
for these functions and automated systems such as ACCCS (Artillery Combat
Command and Control System) have been employed to perform these tasks to
some extent. Since the process of handling ATI is tedious involving handling of
overwhelming data and considerations of various variables simultaneously,
Artificial Intelligence (AI) has the potential to increase the efficacy of artillery in the
battlefield.

India’s Conundrum

India’s efforts at AI R&D, particularly defence R&D, are comparatively


nascent. This lag in AI is surprisingly glaring when viewed with respect to the size
of its economy and defence needs. India’s current AI industry is estimated to be
around $180 million annually. According to one study, India had only about 6 per
cent of the world’s companies and about 29,000 AI professionals in the civilian
sector. In August 2016, Carnegie India published a research paper titled “India and
the Artificial Intelligence Revolution”; the paper highlighted the fact that India
needs to view AI as a critical element of national security in view of the
advancement the world has achieved, and especially in view of neighbour China’s
rapid progress in the field calling India’s entry into the domain as “late”.1

In an ambitious initiative, the Indian Government has started work on


incorporating AI to enhance the operational preparedness of the armed forces in a
significant way that would include unmanned tanks, ships, aerial vehicles and
robotic weaponry. An AI task force has been created to formulate a concrete

1
Pant Atul, Future Warfare & Artificial Intelligence, The Visible Path, IDSA Occasional Paper, August
2018.

PINNACLE 2022 48
Employment of Artificial Intelligence to Enhance Efficacy of Artillery

strategy and framework for employment of AI for national security and defence
needs in the years ahead.

Operational Functions

The armed conflicts during the recent times have ushered in major
changes in war fighting philosophy which now seek to defeat enemy by shattering
his morale and physical cohesion, his ability to fight as an effective whole, rather
than to destroy him physically through incremental attrition. Firepower is going to
play a crucial role, more important than ever before, and contribute substantially in
achieving success in operations. The essence of optimal exploitation of such an
asset of immense firepower lies in certain prerequisites. These are the operational
functions mentioned below which artillery needs to undertake in order to perform
its operational role in the battlefield:-

 Establishment of gap free surveillance grid.

 Handling of ATI (Target acquisition, Analysis and Allotment of


Firepower Asset).

 Passage of engagement orders to all concerned resources &


engagement.

 PSDA and re-engagement, where necessary.

 Operational logistic functions, primarily related to ammunition


management.

Operational Functions and Application of AI

All the above mentioned functions are presently based on human intellect.
The capability of human mind to analyse voluminous data simultaneously,
considering all variables within short duration is limited. Thus, AI presents scope in
each of the above mentioned functions for its employment and enhancing the
overall efficacy of artillery. Present methodologies to perform these functions and
identify the shortcomings which can be overcome with the application of AI.

 Deployment Plan In today’s fluid, crowded battlefield, the


movement and positioning of field artillery units is a very complicated
process that includes location selection, terrain management, movement
planning, the coordination of survey support for firing and target
acquisition. The present methodology for formulating deployment plan
considers the factors like Range, Line To Shoot Down To, Crest
Clearance, Required Gun Density and Survivability against enemy
artillery, air and ground forces. This results in considerable time spent in
recce and selection of gun areas, resolving the issues of availability of
49
PINNACLE 2022
Employment of Artificial Intelligence to Enhance Efficacy of Artillery

route and Battle Space Management. Some form of assistance is provided


by ACCCS which employs GIS based platform to assist in identification of
induction routes and deployment areas. However, it is an automated
system and does not consider all the above mentioned factors in entirety.
AI integrated GIS will allow consideration of various aspects of terrain and
would be able to handle large data obtained from the attributes of the
terrain. These attributes when analaysed with various variables factors for
Artillery deployment will enable evolution of most optimum deployment on
the theatre grid. As the complexity of AI systems matures, AI algorithms
may provide commanders with viable courses of action based on real-time
analysis of the battle-space, which would enable faster adaptation to future
events.2

 Surveillance Grid The deployment of surveillance assets to


obtain ATI involves detailed recce to identify or select the most suitable
location for deploying an asset. Moreover, it requires analyses of
availability, characteristics, capabilities and limitations of surveillance
equipment for generating the deployment plan. Further, sensors deployed
across the battlefield generate a large amount of data pertaining to artillery
targets. This data is required to be analysed in near real time for prompt
engagement. Presently, the procedure is based on human intellect and is
limited to the capability of human mind. An AI based system will enable
quick analyses of the area of interest, identify most probable course of
enemy action and identify the likely locations of the enemy elements. While
BFSRs and EO devices will be ab-initio deployed as per the plan,
thereafter, side stepping based on the progress of the operations and
predicted future locations of the enemy, will be autonomously monitored,
analysed and predicted by AI.

 Target Acquisition and Analysis In the present system, a target


is acquired using manual observation, using Electro Optical (EO) devices
like Thermal Imaging Integrated Observation Post Equipment (TIIOE),
Long Range Reconnaissance and Observation System (LORROS), aerial
platforms, satellite imageries and Weapon Locating Radars (WLRs). Large
number of targets acquired by the sensors need to be analysed in near real
time on various parameters to include location, size, description and
characteristics to identify the most suitable resource for engagement.
Presently various ad-hoc organisations are established at various levels to
carry out this analysis, relying on human intellect. However, the process is
time consuming and thus owing to the limited human capability manifests
in sub optimal results. AI can carry out these tasks autonomously to
analyse, prioritise and allot the most suitable firepower asset. This
aspect when combined with an integrated environment of sensors and
shooters, controlled by AI will shorten the Sensor to Shooter link. These

2
Hoadley D,Lucas N, Artificial Intelligence and National Security, April 2018, Pp 10.
50
PINNACLE 2022
Employment of Artificial Intelligence to Enhance Efficacy of Artillery

type of systems can utilise AI to automatically pursue, distinguish, and


destroy enemy targets and should compose of information collection and
management systems, knowledge based systems, mission implementation
systems, etc.3 For example, Defence Advanced Research Projects
Agency’s (DARPA) Target Recognition and Adaption in Contested
Environments (TRACE) program uses machine learning techniques to
locate and identify targets with the help of Synthetic Aperture Radar (SAR)
images.4

 Engagement Orders and Communication Systems After


analysing the targets, prioritising them and allotting the most suitable
firepower assets, the engagement orders are issued. Presently, some form
of automation has been achieved with ACCCS. However the system is not
AI supported, based on conventional means of communication (Line and
Radio) which causes a delay in passage of engagement orders. The
limited capacity of the media being used, interruptions due to enemy
Electronic Warfare (EW) and the distances involved between the firepower
assets and decision making organisations may render this system
ineffective in an intense battle. AI compatible cognitive radios with dynamic
spectrum management to enhance communications, while pursuing
offensive EW capabilities through the application of machine learning to
learn and rapidly devise countermeasures are the future.5

 Ammunition The assessment of the optimum quantum of


ammunition required to be delivered on the target to achieve the desired
effect is a crucial aspect. With large number of variables like quantity of fire
units, types of equipment, different types of terrain and different types of
ammunition systems, the assessment of the scale of ammunition required
to be fired on the target, cannot be left to the judgement of the observer.
Presently, Ammunition Tables (AMTAB) are used to calculate the
ammunition required to be delivered on the target to achieve the desired
degree of effect. AMTAB allowance is based on limited trials carried out
almost half a century ago. AI based system can carry out analysis of the
target characteristics, lethality of the ammunition and various other
variables involved. AI can be employed to analyse the target
characteristics just by visual acquisition and through photographs.
Thereafter, it can carry out target to weapon matching and based on the
availability of most suitable asset, it can then suggest the quantum of the
ammunition to be delivered on the target to achieve the desired effect.

 Counter Bombardment and Survivability The ability of


producing effective counter bombardment fire depends on the accuracy of

3
Hoadley D,Lucas N, Artificial Intelligence and National Security, April 2018, Pp 10.
4
Ibid .
5
Ibid .
51
PINNACLE 2022
Employment of Artificial Intelligence to Enhance Efficacy of Artillery

location of hostile guns, accuracy of the fire which is being applied for
engagement of these locations and availability of resources for
engagement. The procedure involves detection of hostile gun locations by
WLRs, aerial assets, crater examination and manual acquisition by troops
operating behind enemy lines. The present system is time consuming and
does not ensure full proof integration of all sensors with shooters. Further,
in a dynamic and fluid situation, allotment of resources for this task will be a
challenge. AI based sys will enable quick engagement of enemy gun
areas. Crater examination, shell report and bomb report process can be
automated by utilising AI based system. Just the visual input of the crater
or image of the crater would suffice for the system to decipher the location
of the hostile gun position. This, when combined with the ability to learn
patterns, terrain analysis of own and enemy area, will enable the system to
suggest safe deployment areas to obviate the threat of counter
bombardment by the enemy.6

 Ammunition Replenishment and Transportation To ensure that


no fire power asset is idle at any point of time, it is essential that the supply
of ammunition is continuous. The system presently relies on manual
monitoring and updating for carrying out ammunition replenishment. AI can
be employed to autonomously monitor available stock of ammunition in
various logistics echelons. AI can lower transportation costs and reduce
human operational efforts. It can also enable military fleets to detect
anomalies and predict component failures.

Conclusion

The dynamics of the present and future battlefield demand near real time
combat intelligence and engagement, which existing systems cannot provide. The
proliferation of sensors and fluid battlefield situation will generate enormous
volume of information, which will be humanly impossible to process. It is therefore
imperative to develop a fully automated, autonomous and intelligent system, which
will receive information from directly interfaced system, carry out target
identification and situational assessment, present a cohesive picture and nominate
the firepower resources for near real time engagement of targets.

6
Thesis on Artillery Survivability Model performed by MOVES Institute.
52
PINNACLE 2022
ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING:
THE WAY AHEAD FOR FUTURISTIC IMAGERY
INTERPRETATION
- Captain Shreya Sood

“Artificial Intelligence is in a „Golden Age‟ and solving problems that were


once in the realm of science fiction.”
- Jeff Bezos, Founder and Chairman, Amazon

Introduction

Artificial Intelligence (AI) and its subset, Machine Learning (ML) is a


trailblazing technology that can mimic the capabilities of the human brain in
machines. It is the main driver of emerging technologies like big data, robotics,
and Internet of Military Things (IoMT), and will continue to act as a technological
innovator for the foreseeable future.

Interpretation of space and satellite images is one of the sectors being


changed by AI and ML. The ever-growing number of satellites in space has
resulted in massive amounts of data received by ground stations on a regular
basis, increasing the requirement for skilled manpower, training, and technological
resources to process it. The use of AI and ML in this field will be a viable option
as it reduces processing time and effort while enhancing productivity and error-
free outputs. Geospatial tools enable the prediction, monitoring, and countering of
such threats, as well as the planning and support of field operations.

AI and ML in Satellite Imagery Interpretation

Every day, millions of photos are shot in space and returned to Earth by
ever-increasing number of satellites. Satellite imaging enables unprecedented
degrees of remote surveillance of an adversary's locations, movements,
concentration, training drills, naval manoeuvres, damage assessment, and so on.
They assist military observers and commanders in detecting changes in hostile
terrain across a large area. Satellites are now transmitting significantly more data
for processing and analysis. As a result, AI might be utilised to automate the study
of this data from various sources and alert analysts to unusual events that require
their expertise.

In light of this, one of the most significant advancements in Imagery


Interpretation (IMINT) is the use of software and AI algorithms to analyse satellite
imagery. Despite the ever-growing volume of photos, the algorithm now in use
allows an interpreter to detect, recognise, and identify aeroplanes, vehicles, and
ships. In addition to assisting experts in their task of looking for information and
hints on photos, this technology also enables them to enhance the study of
activities of previously chosen geographic locations. It helps to lighten the load on

PINNACLE 2022 53
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

human analysts so they can handle the growing amount of imagery data being
collected.

Apart from better clarity and visualization, ML lets us interpret images in


many ways and from different vantage points. The scope of using ML in satellite
imagery is not restricted to a narrow field but could be used for modelling,
separating images or extracting useful information. The process is complex but the
final results are worth the long process. Some of the other cutting-edge benefits of
application of AI and ML to satellite imagery have been enumerated below:-

 Numerous applications that demand a significant amount of spectral


and spatial data for pattern recognition benefit from ML and remote
sensing technologies.

 ML combines algorithms using Computer Vision (CV), computer


systems, and deep learning techniques to collect and classify features
quickly and accurately.

 ML and CV, can help the end-user to process the data collected to
find solutions from remote sensing data. It can also improve the overall
accuracy of the data classification and enhance the reliability and
assessment – IMINT analysis.

 To overcome the limitations of remote sensing data, ML techniques


through neural networks can enhance the analysis of broad areas in order
to classify objects, identify temporal change, data fusion, cloud removal,
and spectrum analysis from satellites or aerial pictures.

 For areas with complex feature distribution, such as in the


applications of environmental monitoring and management, mineral
mapping, agriculture, disaster management, climate change, and wildlife
conservation, images collected by satellites or Unmanned Aerial Vehicles
(UAV) using ML techniques can provide near real-time reports.

Applications of AI and ML in Satellite Data Modern AI and ML


techniques, particularly deep learning, have made it simpler to perform tasks such
as object identification, item counting, semantic segmentation, and general picture
classification. Following are a few examples of AI and ML applications in satellite
imagery: -

 Object Detection Using ML Object detection is a Computer


Vision technique for locating instances of objects in images or videos.
Object detection algorithms typically leverage ML or deep learning to
produce meaningful results. This happens when Computer Vision
techniques are used to identify various features of an image, such as the
colour histogram or edges, to identify groups of pixels that may belong to
54
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

an object or any changes within the same. These features are then fed into
a regression model that predicts the location of the object along with its
label. Object detection is not just a requirement for contemporary combat
but also a key factor in the development of early strategic warning systems.
A large dataset is needed in order to train an AI model to recognise military
artefacts. There must be a significant number of photos with labels in the
dataset. Analysts can automate detection and analysis of activity-based
data by combining electro-optical satellite imagery with advanced
algorithms to detect and flag objects with high accuracy to include trucks,
tanks, ships, and aircraft.

Figure 1: Automated Algorithms Detect Airplanes in High-Resolution


Satellite Imagery

 Change Detection Algorithms It is defined as the process of


identifying differences in the state of an object or phenomenon by
observing it at different times.

 This process is usually applied to detect changes in earth


surface at two or more times. The primary source of data is
geographic and is usually in digital format such as satellite imagery,
analog format (aerial photos), or vector format (feature maps).
Ancillary data (historical, economic, etc) can also be used.

 Actual information comparing buildings, road segments,


troop concentration, new construction, concentration of weapon
systems, mobilisation, etc is extremely important from a military
point of view. Technically, the goal of a change detection algorithm
is to create a map in which changed areas are separated from
unchanged ones, and the value of AI is in replacing visual
interpretation with automatic or machine interpretation, which would
increase the amount of data being processed while simultaneously
decreasing the amount of human intervention or probability of error.
55
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

True Orthophoto Date 1 True Orthophoto Date 2

Figure 2: Time-Series Analysis and Change Detection Algorithm


Demonstration. Changes in Buildings and Roads Highlighted Red
(Before) to Green (After)

 Terrain Analysis By combining stereo satellite imagery


and terrain elevation databases, it is feasible to create true-color 3D terrain
visualization models of any location. Detailed knowledge of the terrain is a
key factor for successful military operations. Stereo satellite map data is a
source of reliable, regularly updated information that gives a realistic and
objective view of the terrain which is indispensable for mission planning.

 CV Techniques The fundamental use of satellite data is to enable


CV to detect diverse objects. CV is a branch of AI that allows machines
and programs to extract useful data from virtual photos, videos, and other
sensory inputs. CV can also participate in executing certain measures to
make predictions based on that data. Municipalities, government bodies,
emergency crews, the armed forces, and other civil authorities require
reliable information on structures, route sections, and urban area
boundaries.

 Automated Response System AI onboard satellites can enable


the collection and transmission of information in near real time. On-orbit,
ML algorithms may be used to detect anomalies and initiate responses.
Tasking another imaging satellite to collect pictures at a specific timestamp
or coordinating a co-collection activity of several data kinds in the same
region of interest are examples of autonomous replies. Inter-satellite
communications links would be used to send tasking directives and data
acquired. Algorithms can analyse, categorise, or combine massive
56
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

amounts of data on-orbit with enough on-board processing gear on each


satellite to offer insights straight to end users when they need them.

 Estimation of Global Oil Inventory Employing the basics of


interpretation and trigonometry, this is achieved by interpreting shadows.
As most of the fuel tanks have a floating top, it is possible to calculate its
volume and the amount of oil in it, utilizing simple trigonometry. Also,
knowledge about the locations of all such oil tanks (extracted from other
satellite imagery or gathered from different database) can assist in
area/region-wise calculation of global oil inventories.

Figure 3: Measuring Oil Inventory by Interpreting Shadows

 Monitoring National Activities Another application which can be


said to combine all the other applications given above is monitoring the
economic and military activities of various countries across the globe. If
independent satellite-powered measures of cities’ development, such
as car density, construction rates, electricity consumption through nighttime
illumination, import and export operations through ship count and sizes,
and aircraft count in airports, combined with traditional surveys, statistics,
and media data are indexed/catalogued, it can generate a comprehensive
report on the economic, financial, and social health of the country or region
which is otherwise off-limits.

57
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

Figure 4: Construction rates monitoring via shadow detection

 Creation and Use of Synthetic Data To build an effective AI


system, engineers need to feed massive amounts of training data to ML
algorithms, images of the type of objects they want them to find, so they
can automatically find those objects when presented with new images. To
compensate for the shortage of such data, real world data is often
supplemented with synthetic data, which is artificially created but designed
to look like the real-world data the machine is being built to work with. Most
AI algorithms are trained on both real-world images and synthetic images.

 Quantum of Imagery Data There is a large amount of imagery


and other types of data transmitted by satellites to reception stations on
Earth. It can neither be looked at nor exploited in its entirety by traditional
means. Hence, actions like counting objects in imagery, monitoring
different places, etc. it can be processed by AI/ML. The military and
intelligence setups of many countries, such as the USA, Israel, Russia,
France, etc., are keen on using AI tools to sift through the vast torrent of
data created by an ever-increasing number of sensors and pick out the
most important information for human analysts.

AI Based Software in Satellite Imagery Interpretation


Some tools and software in vogue or under development for both military
and civil use have been covered in the succeeding paragraphs. Such tools allow
armies to increase their surveillance and intelligence capabilities by processing the
growing volume of data coming from new sensors and by relieving experts of
repetitive tasks. The army thus has greater efficiency in situation assessment and
decision-making. Analyzing images is a crucial issue, and a well-coded AI can
generate alerts as soon as predefined parameters are observed. As a result, it is
not surprising that security laws are being enforced zealously to prevent pilferage
58
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

or unintentional sharing.

 Planet Labs International aerospace and data analytics firm


Planet of the USA has a dedicated solution for ML called Planet Analytics.
This uses ML algorithms for processing daily satellite imagery, detecting
and classifying objects, locating topographic and geographical features,
and consistently monitoring even the most infinitesimal change over time.
The information feed is seamlessly integrated into the workflows and
provides insights on virtually any location on the planet.

 Descartes Lab Satellite imagery refining start-up Descartes Lab


based in USA has a cloud-based platform that applies ML forecasting
models to petabytes of satellite imagery that is drawn from a number of
sources.

 Google One of the pioneers utilising the prowess of ML in satellite


photography is Google. Google has released a tool named PlaNet that can
pinpoint the precise geographic location of any image taken anywhere on
the planet. Convolutional neural networks and mapping technologies are
the foundation of the PlaNet project, which harnesses the potential of ML,
offering knowledge that is priceless and unmatched in both qualitative and
quantitative ways.

 Preligens The French company Preligens employs a tool that


compiles data (commercial and open source) from a variety of sensors
(satellite images, air and ship transponder data, infrared imagery, social
networks, etc.) and notifies analysts in the event of unusual changes in the
situation at a specific theatre of operations. In particular, it can track the
development of a critical site and automatically detect armoured vehicles,
aircraft, or ships. When a nuclear submarine departs from a military port or
when an unusual concentration of armour is found, the AI can notify a
human analyst.

 Project Maven It is a United States Department of Defense


program to create an AI-powered investigation system for UAVs to prepare
predictive analytics of drone footage.

 OneAtlas Project OneAtlas is prepared by Airbus. The software


provides a collaborative environment to easily access very high-resolution
imagery, performs large-scale image processing and extracts industry
specific insights. The services include infrastructure change detection,
vehicle detection & counting and will soon cover aircraft detection and land
use change detection as well.

 Raster Vision Raster Vision is an open source framework for


Python developers building computer vision models on satellite, aerial, and
59
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

other large imagery sets (including oblique drone imagery). There is built-in
support for chip classification, object detection and semantic segmentation.

 Pulse Satellite Pulse Satellite is a tool to analyze satellite imagery


assisted by neural networks. It has been developed by United Nations
Satellite Centre.

Challenges and Concerns

The challenges associated with geospatial data are regarding its


processing and analysis. Some of these challenges have been enumerated
below:-

 Size The size of objects in satellite imagery is usually very small


(~20 pixels), yet the size of the input images is vast (often hundreds of
megapixels). The algorithms outlined above are not tuned to detect minute
objects in huge images and frequently perform poorly when applied to
Earth observation data due to a relative lack of training data. Although this
field of study is still relatively new, adapting these methodologies to the
many scales and objects of interest in satellite imagery shows considerable
promise.

 Amount of Data and Storage According to research, the


expanding fleet of imaging satellites transmits back to Earth over
80 gigabytes of data daily. This information is unprocessed and binary in
nature. Professionals who access the data must be aware of their search
criteria. Additionally, the cost of storing such data is substantial. In order to
distil the layers of data included in the images into coherent information
that is helpful to other academics, politicians, or funding organisations,
additional processing power and enhanced human experience are needed.

 Normalisation of Images Another complexity is that all images in


the dataset should be normalised to be ready for input into the change
detection process. Building change detection solutions requires
competence in ML and strong experience with remote sensing data.

 Automatic Target Detection Due to the target object's varied


size, orientation, scale, and background, this is a difficult task that largely
relies on pre-existing data that will be used in AL/ML techniques. In
addition, very clear and concise attributes are present in the edge
information of objects in satellite photography. AI systems can
automatically detect changes in satellite photos while taking into account
the image's radiometric properties. However, accurate interpretation of the
change as well as accurate identification of targets of military relevance
requires human participation. The same will only reveal the locations of the
temporal changes in the image. A well-defined and extremely particular
60
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

data set must be developed and supplied to the machine in order to


minimise or eliminate human interaction in interpretation and identification
using AI. Further, such military data will also be strictly classified. This is a
major problem because creating a repository of such a classified data
pertaining to military information of foreign or hostile military forces is
extremely difficult and time-consuming. Such data will be dynamic in
nature, and therefore inaccuracies in the same is a probability which
cannot be overlooked. Even if acquired, it will be sensitive and classified
due to its inherent nature, creating additional concerns about its handling
and dissemination.

 Type of File Most popular deep learning architectures are not


designed for imagery that is often a gigabyte or larger, may contain over a
dozen channels (most of which are not in the visible spectrum), and is
stored in spatially referenced file formats like GeoTIFF and JPEG2000.
Hence, for professionals interested in applying these techniques to satellite
imagery, there remain many obstacles to even basic workflows.

Prognosis

The Armed Forces cannot remain oblivious to such technological


advancements in the field of Satellite IMINT and therefore will have to devise
methodologies to keep pace with such cutting-edge technologies. The same would
entail a combined and synergized effort of all concerned departments and
stakeholders.

A holistic approach towards the creation of capability and imbibing of such


cutting-edge technology by combining the efforts of all concerned lead agencies
dealing with the subject is the need of the hour. While attempts are being made to
develop the same, there is overlap between the activities of several entities
working on related projects. All interested parties must build a shared roadmap for
creating AI and ML-based solutions for imagery interpretation in the armed forces.
This would entail inter-agency or departmental synergy and involvement of subject
experts from both within and outside the organization. Creation of a pool of subject
experts within the organisation would therefore assume priority.

National institutions such as ISRO, the DRDO's Centre for Artificial


Intelligence and Robotics (CAIR), and others should be approached to take the
lead in this regard. Involvement or inclusion of select indigenous private players or
startups in the subject industry could also be considered, with due security vetting.
The National Geo Int Framework, designed by the NTRO as a tool to assist in
decision making and analysis by concerned stakeholders at the national level, on
aspects concerning National Security, is expected to roll out by October 2022. Ibid
tool will also result in creation of a repository of data from all sources of
intelligence, especially data of military significance. Such data will be highly
classified with the necessary security protocols for access. Such information will
61
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

be extremely useful to all Armed Forces, including the Defence Space Agency
(DSA). Utilisation of above for creating a framework for introduction of AI and ML
should be thought about and planned by the NTRO in consultation with DSA and
the three services.

Bibliography

Publications/ Books

1. Du, E., Ives, R., van Nevel, A. et al. Advanced Image Processing for Defense and Security
Applications. EURASIP J. Adv. Signal Process. 2010, 432972 (2011).

2. Military Intelligence Satellite, Dr Manish Chandra, Surendra Publications, 2018, New Delhi.

3. Artificial Intelligence: A Guide for Thinking Humans, Melanie Mitchell, Pelican Publications,
2019, UK.

4. A fast military object recognition using extreme learning approach on CNN by International
Journal of Advanced Computer Science (IJACSA) and Applications, Vol. 11, No. 12, 2020.

5. Artificial Intelligence and the Future of Power: 5 Battlegrounds, Rajiv Malhotra, Rupa
Publication, 2021, New Delhi.

Web Links

1. https://www.geospatialworld.net/blogs/machine-learning-in-satellite-imagery/.

2. https://earthobservatory.nasa.gov/features/OrbitsCatalog.

3. https://medium.com/vsinghbisen/ai-applications-for-satellite-imagery-or-satellite-images-
dataset-3b1a2499c5e5.

4. https://techfastly.com/satellite-imagery-and-ai/.

5. https://emerj.com/ai-sector-overviews/ai-applications-for-satellite-imagery-and-data/.

6. https://www.coe.int/en/web/artificial-intelligence/history-of-ai.

7. https://opendatascience.com/the-history-and-future-of-ai-with-michael-i-jordan/.

8. https://www.geospatialworld.net/blogs/difference-between-ai%EF%BB%BF-machine-
learning-and-deep-learning/.

9. https://www.esa.int/Enabling_Support/Preparing_for_the_Future/Discovery_and_Preparatio
n /Artificial_intelligence_in_space.

10. https://concisesoftware.com/history-of-machine-learning/.

11. https://insights.regenesys.net/how-is-artificial-intelligence-transforming-business/.

12. https://web.archive.org/web/20130719163905/http://unpan1.un.org/intradoc/groups/public/d
ocuments/apcity/unpan002006.htm.

62
PINNACLE 2022
Artificial Intelligence and Machine Learning: The Way Ahead for Futuristic Imagery Interpretation

13. https://cloud.withgoogle.com/build/data-analytics/explore-history-machine-learning/.

14. https://www.ibm.com/in-en/cloud/learn/what-is-artificial-intelligence.

15. https://www.sas.com/en_us/insights/analytics/what-is-artificial-intelligence.html#industries.

16. https://insights.regenesys.net/ai-and-machine-learning-the-relationship-explained/.

17. https://concisesoftware.com/history-of-machine-learning/.

18. https://insights.regenesys.net/how-is-artificial-intelligence-transforming-business/.

19. https://www.fritz.ai/object-detection/.

20. https://www.pixxel.space/technology.

21. https://www.mdpi.com/2079-9292/7/10/216/htm.

22. https://www.popularmechanics.com/military/research/a37014065/how-to-spy-on-military-
bases-use-satellite-imagery/.

23. https://wsimag.com/science-and-technology/64215-artificial-intelligence-has-changed-our-
world.

24. https://www.weforum.org/agenda/2016/10/top-10-ethical-issues-in-artificial-intelligence/.

25. https://whatis.techtarget.com/definition/AI-code-of-ethics.

26. https://www.brookings.edu/research/how-to-address-ai-ethical-dilemmas/.

27. https://orbitalinsight.com/geospatial-solutions/defense-intelligence.

28. http://interactive.satellitetoday.com/via/october-2020/10-ways-ai-is-making-a-difference-in-
the-satellite-industry/.

29. https://arxiv.org/pdf/1812.05815.pdf.

63
PINNACLE 2022
ARMY DESIGN BUREAU - CONNECTING CAPABILITIES IN
ARTIFICIAL INTELLIGENCE FOR NEW AGE WARFARE
- Colonel Amandeep Singh Bains

“Victory will smile upon those who anticipate changes in the character of
war, not upon those who wait to adapt themselves after changes occur.”
- General Giulio Douhet, The Command of the Air (1921)

Introduction

In recent times, the impact of technology has assumed an altogether


new and disruptive dimension. Science & Technology has a central role in the
defence eco-system as it drives strategic national capability and is central to
countering emerging threats. Our modernisation strategy for developing Artificial
Intelligence (AI) needs to focus on information and data at the core. As we
modernise, we will embrace information-centric technologies, recognising that it
will be the application of a combination of technologies like processing power,
connectivity, AI, automation, robotics, autonomy, and quantum that will
achieve the disruptive effect we need. 1

In the realm of National Security, AI presents huge military opportunities


and its associated challenges. As conflicts increasingly leverage technologies
like AI and Big Data, military relevance will rest heavily on the paradigm where
technology will achieve strategic objectives, while preventing the adversary from
doing the same. It is important to understand how AI will develop over time in New
Age Warfare, its impact on operations and capability and their applications on
military concepts and organisational structures.

AI Enabled Battlefield
The underlying technologies
Military capability is not only themselves (the internal
dependent on fielding weapon systems combustion engine, radio
with superior technology. History has communications, radar etc.) as well
shown that improving military
as the new military systems to
effectiveness requires organizational and
doctrinal innovation. For example, which they gave birth (airplanes,
German Blitzkrieg was a military concept tanks, amphibious landing craft,
that combined organizational and aircraft carriers, radar and so forth),
doctrinal innovations.2 One still had to formed only a part of the
integrate advanced weapons systems innovations.
with appropriate tactics, operational

1
https://www.gov.uk/government/speeches/chief-of-the-defence-staff.
2
Emerging Technologies and Military Capability Dr Andrew D. James.
PINNACLE 2022 64
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

concepts and doctrines in order to realize the full potential of new ways of fighting
(see box)3. This is where the challenge to integrate AI into systems and war
fighting lies.

Technology has come to dominate


Emerging technologies will shift the
the discourse of military strategists in the
aftermath of the recent conflicts in Israel, balance between quality and
Armenia-Azerbaijan and Russia- quantity, as well as between
Ukraine war. The Israeli military is offense and defense. For
calling „Operation Guardian of the example, low-cost drone swarms
Walls‟, the first AI War4, where AI was a could overwhelm defensive
key component and force multiplier. systems, providing a greater
Israeli Defence Forces used advanced advantage to the attacker, while
program „Alchemist‟, „Gospel‟ and Counter-Drone systems could
„Depth of Wisdom‟5 during the fighting easily neutralize such attacks, and
where „Alchemist‟ used AI to alert troops favor the defender.
of possible attacks and „Gospel‟
generated recommendations for aerial targets. It is believed that AI helped
shorten the duration of the fighting, having been effective and quick in
identifying targets.

At the same time, the recent events in Saudi Arabia, the Azerbaijan-
Armenian conflict over the contested Nagorno-Karabakh region, and Ukraine
are reminders of the paradigm and showed how UAVs/Drones can dominate
ground operations and inflict crippling losses. Future wars will also involve low
technology, which is easy to obtain but difficult to defeat. Since World War II,
high-tech militaries have been consistently
thwarted by low-tech opponents. Thus, there is 4 IR Technologies
a need to maintain a balance while infusing
technology into our inventory and doctrines.  AI
 Robotics
 IoT and Cloud Computing
AI - Enabling Fourth Industrial Revolution  Quantum Computing
Technologies  Blockchain
 Big-data Analytics
AI is playing a predominant role in  Novel materials
connecting Fourth Industrial Revolution (4IR)6  Additive manufacturing
The development of 4IR-enabled military  Bio Technology
 Smart Sensors
capabilities raises a critical analysis in the  VR & AR
difference between four important concepts  Energy Storage
that sit roughly on a continuum ranging from  Space Technologies
science to technology to capability to effect.

3
Barry Watts and Williamson Murray in Military Innovation in the Interwar Period (In experimentation:
The Road to Discovery - Tom Greenwood and Jim Greer, 2018).
4
Anna Ahronheim in ‘The Jerusalem Post” on 27 May 21.
5
Younis Dar in ‘Eurasian Times’ on 29 May 21.
6
SIPRI Report on Emerging Military and Security Technologies.
65
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

Moving from technology to capability requires a range of non-technological


innovations while moving from technology to effect is largely a matter of effective
implementation.7 We need to appreciate that ‘Uncertainty’ is a key characteristic
of technological change and the process from the emergence of a technology to its
having an impact on military capability is a long one. 8

Figure 1: Continuum of Niche Military Technology Capability


Development

Army Design Bureau: Focus Areas for AI Capability Development

Future combat outcomes will rest on the development and deployment of


AI-based military capabilities and applications. Today, when we are looking at the
contours of future warfare, the development of weaponry and other systems is
being done keeping in mind the predominant play of AI.

The emerging battlefield environment will be dynamic, multi-dimensional,


and technologically intensive, characterised by enhanced battlefield transparency,
real-time connectivity, and extreme lethality. In order to maintain a technological
advantage across key capability areas, certain AI use cases that are being
developed by the Army Design Bureau through the Indian industry, start-ups,
and academia are highlighted.

Figure 2: AI Domains

7
Atlantic Council Report on Emerging Technologies.
8
Emerging Technologies and Military Capability Dr Andrew D. James.
66
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

 Natural language Processing


Development of bi-directional English to
Mandarin devices, to develop ruggedised
„Wearable Translators‟ which are custom
designed to work in the edge.

 Autonomous Systems The use of AI


for autonomous systems is a priority and the
Swarm Drone demo witnessed during the
Army Day celebrations in 2021 was only a
beginning. The applications for manned
unmanned teaming, Unmanned Ground
Vehicles (UGVs) for Intelligence,
Surveillance and Reconnaissance (ISR),
Combat and Support applications are being
developed.

 ISR Satellite and UAV imagery


analysis has huge potential and projects which
will enable automatic identification and
change detection to enable event alerts, will
be extremely useful in defending our borders.

 Threat Modelling Use cases like aerial threats from drones and
suspicious vehicle detection in Counter Insurgency are a focus area for
development.

 Logistics AI has tremendous utility in


the field of military logistics. It is a priority area
for capability development in Supply Chain
Management and Inventory Control.

 Health AI can play a predominant role in


health care. A unique use case which focuses on
identifying early indicators in the development
High Altitude Illness and will help in reduction
in morbidity and mortality of troops is
underway.

Given the reality of short technological cycles, the rapid pace of


obsolescence, and rising costs, it has become imperative today to align our AI
technological aspirations with the civilian domain. The Army Design Bureau is
matching the Indian Army’s aspirations with what the industry can
realistically deliver and thereafter progressing AI use cases through multiple
67
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

routes, including initiatives like collaborating Routes for AI Development


with Defence Public Sector Undertakings
(DPSUs) like BEL in co-development of AI  MAKE II
use cases. Today, in all future procurements  iDeX (Def Innovation Org)
and upgrades, we are considering the  Army Tech Board/R&D
incorporation of AI for all Qualitative  Tech Devp Fund (DRDO)
Requirements (QRs).  Capital Procurement
 Collaboration with DPSUs
Data - Strategic Asset for AI Capability

The ability to undertake Algorithmic Warfare heavily rests on computing


technology, which is dependent on Processing Power, Big Data and Cloud
Technology.9 Our modernisation strategy needs to focus on information with
data at the core. As we modernise, we will embrace information-centric
technologies, recognising that it will be the application of combinations of
technologies like processing power, connectivity, machine learning and artificial
intelligence, automation, robotics, autonomy and quantum computing that will
achieve the disruptive effect we need.
The Indian Army is graduating from an ‘era of log books’ to standard
data management framework. The need of the hour is to improvise on the fly
to breed a data centric culture, which will allow us to provide training data
for AI Engines. Today, capability exists to derive a trained AI model from general
data, which can be „rebuild and retrained‟ by methodologies like „transfer
learning‟ and techniques that convert unstructured data to the required
format to train AI engines.

AI systems will be the future enablers for the Armed Forces to operate in
the future Multi Domain Operations where AI-infused autonomous weapon
systems will become Force Multipliers and be the fulcrum around which Big
Data will be turned into actionable intelligence and a decision advantage. We
in the Army realise that AI needs data for the development of intelligent machines,
and we need to breed a culture in which we recognise and value data as a
strategic asset!

Army Design Bureau - Facilitating Experimentation and Risk

The 21st Century is a „risk-on’ environment and does not reward timidity -
if we remain cautious then we carry significant risks. AI and Big Data will have a
disruptive impact on war fighting and we have to pay adequate emphasis on AI
technologies that have dual use and are being driven by commercial entities and
innovations. However, predicting changes will be challenging, so we will have
to take risk and accept some failures.

9
‘Algorithmic Warfare - Applying AI to War fighting’ - Peter Layton.
68
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

Figure 3: Hype Cycle - AI Innovations10

The above graph figure shows a pattern that arises with each innovation/
technology. In the Army as we integrate AI, we will also have our breakthroughs in
AI technology, buzz around the potential operational expectations, failure &
frustration when use cases are delayed, experience benefits of commitment in few
AI use cases and finally Capacity Enhancement and Capability Development
in critical AI domains.

Technological development requires experimentation and failure, and


there is a need to incorporate a greater appetite for risk-taking into our
innovation pipeline. We have to realise that technological development requires
experimentation, and in order to leverage technological advantage, there is a need
to incorporate an appetite for risk-taking to support start-ups. There is substantial
progress in R&D in niche areas along parallel routes so as not to lose time if
one fails or stumbles, which would put the clock back a few years.

Army Design Bureau - Connecting Capabilities in AI

The Indian Army has been advocating the need for increased self-reliance
in the defence sector as only indigenous technologies will be available during
conflicts. In order to take advantage of new government policies to foster a robust
defence industry in emerging technologies/systems and realise our Atmanirbhar
Bharat Vision, the Army Design Bureau is at the forefront of all initiatives.

Innovation is a key tenet to maintain operational advantage and to make


defence an attractive innovation partner, the Army Design Bureau is finding ways
to share risk and reward, provide clarity on intentions and make contractual
processes less daunting in order to encourage and incentivize technology/

10
Steinert, Martin. "Scrutinizing Gartner's hype cycle approach". Research Gate. IEEE Xplore,
September 2021.
69
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

innovations. The Indian Army is striving to ensure that we do not let go of


potentially impactful advancements in AI languishing in the ‘Valley of Death’
between invention and adoption.

The Indian Army established the Army Design Bureau as a single point
of contact with all innovators, industrial houses, DRDO and budding
entrepreneurs. The ADB has initiated a focused Outreach Program in AI
applications with the Industry and Academia and significant progress has been
made. Initiatives in getting Industry, Academia, R&D establishments & Start
Ups together have started bearing fruits towards rapidly advancing technological
needs. ADB is also looking at establishing mechanisms for technical cooperation
and co-development of AI Technology through strategic alliances. The time is
now apt where there is a need to augment the private sector with enabling
provisions and facilities to make them the starting blocks of R&D in AI.

Tech
Support

Industry
Global
Connect

Army
Design
Bureau
MoD Academia

Innovation DRDO

Figure 4: Army Design Bureau: Connecting Capabilities in AI

The rate of technological change, places a premium on agility and


responsiveness of the acquisition system and the need of the hour is to
gravitate towards an open model of defence innovation for faster infusion of
technology into the Army. We are also looking to have a dedicated capability
that will crystal gaze and identify how AI military applications will impact
warfare and which AI domain is worth investing in. At the same time, we also
need to ensure that if the desired military capability is not being met within our
defined timelines, resources and budget then it is time to rethink that
technology/equipment concept.

As the R&D interface of the Indian Army, the Bureau is creating an


innovation culture in AI for startups through steps like facilitating trials in field
conditions, establishing Cells at Indian Institute of Technologies (IITs) with
70
PINNACLE 2022
Army Design Bureau-Connecting Capabilities in Artificial Intelligence for New Age Warfare

Service Innovators posted, reaching out at various forums like ARTECH,


DefExpo/AeroIndia etc. This all has been achieved by working with an open mind
with the academia, industry, in-house innovators, investors etc and having a
critical, decisive, failure absorbing, agile and responsive technology development
model.

Conclusion

The Chief of Army Staff has often highlighted the impact of disruptive
technologies in warfare and emphasised that our current modernisation drive is
focused on upgradations with available indigenous technologies.11 Today, the
domain of ‘Non-Contact Warfare’ is as important as ‘Contact Warfare’; and
the virtual fight is assuming as much salience as the physical fight - that is where
AI will play a big role.

The envisaged contours and trends of future warfare in the era of AI


are evolving and we need to be agile to adapt to it. Timely infusion of technologies
like AI, Machine Learning and Big Data Analytics is crucial, so that we are not
left behind the technological curve. The Army Design Bureau is focused on
developing technologies and capabilities in AI, by imbibing an innovation
culture for harnessing emerging technologies and building an effective Services-
Industry-Academia interface.

11
https://pib.gov.in/PressRelease, August 20.
71
PINNACLE 2022
ARTIFICIAL INTELLIGENCE IN MODERN WARFARE:
DRONE DETECTION USING MACHINE LEARNING
- Captain SBS Bhullar

Introduction

There have been numerous incidents of hostile or terrorist strikes using


drones for surveillance. These are just a few of the numerous potential
applications for these devices, so it's important to identify and monitor drones in
locations where potential assaults can take place. These aircrafts have the
capacity to fly both very low and very high1. These platforms provide a serious
threat since they can move stealthily and unobserved. The drones consequently
become the perfect tool for compromising security, putting lives in danger, inflicting
damage or unauthorized entry into one's own area, as well as monitoring military
activity.

Factors of Drone Detection

 Appearance and Operation of Drones Some drones include a


light device on the lower half of their bodies, which can make detection
more successful, especially in low-light situations. The way that a drone
operates is another characteristic that sets it apart. The movement's goal is
to take the drones from its stabilised position to its final destination, where
it will be stabilised once more. The movement of drones is recognised for
its predictability. The drone does not make any of these motions, in
contrast to a bird gliding or flapping its wings.

 Basic Approaches Two categories-visual and acoustic can be


used to categorise the fundamental methods of drone detection. Visual
detection uses camera systems which capture a two-dimensional image on
how the drone appears or moves, or perhaps on both. Acoustic
approaches employ microphones2 to capture acoustic signatures even if
they are not visible.3 Combining these two strategies will result in a hybrid

1
Norouzi Ghazbi, S.; Aghli, Y.; Alimohammadi, M.; Akbari, A.A. Quadrotors Unmanned Aerial
Vehicles: A Review. Int. J. Smart Sens. Intell. Syst. 2016, 9, 309–333.
2
Case, E.E.; Zelnio, A.M.; Rigling, B.D. Low-cost acoustic array for small DRONES detection and
tracking. In Proceedings of the 2008 IEEE National Aerospace Electronics Conference, Dayton, OH,
USA, 16–18 July 2008.
3
Busset, J.; Perrodin, F.;Wellig, P.; Ott, B.; Heutschi, K.; Rühl, T.; Nussbaumer, T. Detection and
Tracking of Drones Using Advanced Acoustic Cameras. In Unmanned/Unattended Sensors and
Sensor Networks XI; andAdvanced Free-Space Optical Communication Techniques and
Applications; International Society for Optics and Photonics: Bellingham, WA, USA, 2015; Volume
9647.
PINNACLE 2022 72
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

detection system.4 For instance, a radio-frequency jamming device, radio-


frequency sensor, and video are all necessary components of an anti-
drone system.

 Object Characteristics Algorithms for object recognition and


classification are based on the properties of the objects being taken into
account. This implies that the traits are determined by their behaviour and
movement patterns in addition to how they look. The challenge of
identifying drones in the sky centres on both the drones themselves and
other items that could appear in the sensing area. One of a drones most
glaring visual features is its form. From a tricopter to the octocopter, every
form of drone resembles the others in their respective categories. A
quadcopter is shaped like a square, a tricopter like an equilateral triangle,
etc. Each drone also consists of a hard structure with recognisable visual
traits. Depending on the number of propellers, the control board is often
positioned in the centre of the structure and has three to eight arms. Each
arm has a motor at the end, which is attached to the propeller. The most
fundamental drone appearance pattern is this one, and it may be used on
practically every drone.

 Long Range Surveillance and Armed Drones There is another


kind of drone that has a different size, weight, and top speed than multi-
rotor models. They are known as fixed-wing drones. Due to their
characteristics, they are commonly used in applications like meteorology,
quality inspection, and environmental and area mapping (with the capacity
to further analyse and generate three-dimensional data). Additionally,
important benefits include their flight safety and operational range.

 Onboard Components The inclusion of a variety of distinctive


components is another way that drones are differentiated. Some drones,
for instance, provide protection to the machine by enclosing the entire
circumference in a ring. On the other hand, some components of other
drones kinds are entirely accessible. On some drones models, an
additional covering ring that is situated close to the machine propellers may
be used. A camera is a standard feature on most drones. The drone's body
can be modified to conceal the camera, minimising its external visibility.
However, on drones with higher-quality cameras, the camera is mounted
on the underside of the drones, where it is clearly visible. The existence of
additional electronics helps to distinguish some drone types.

4
Shi, X.; Yang, C.; Xie,W.; Liang, C.; Shi, Z.; Chen, J. Anti-drone system with multiple surveillance
technologies:Architecture, implementation, and challenges. IEEE Commun. Mag. 2018, 56, 68–74.
[CrossRef].
73
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

Existing Methods of Image Processing

 Contour Searching Creating a curve that encloses the objects in


the image is the main idea behind employing contours in image
processing. This object bounding method can only be used successfully if
picture preprocessing techniques like image smoothing and morphological
procedures are used. The image must only be split into positive and
negative regions,5 whose edges can be visualised as delimited objects, in
order to use contour searching. A contour is a collection of points that
depicts a curve in a picture. These curves are shown as sequences, with
each record storing information for the subsequent point on the curve. Due
to the way it is structured, the contour searching function might produce a
"contour tree." Thus, it can be determined which contours are child
contours and which contours are root contours.6

Figure 1: Contour Searching

 Selective Searching An algorithm called selective searching is


one of the most successful ways to discover sub-regions in an image that
contain an object. This method is based on three key assumptions7:-

 Capturing All Possible Scales in the Image Selective


searching employs a hierarchical approach to account for all
possible object scales.

 Diversification Selective searching does not utilise a


uniform technique for a subregion search since objects in the

5
Bradski, G.; Kaehler, A. Learning OpenCV: Computer Vision with The OpenCV Library; O’Reilly
Media, Inc.:Newton, MA, USA, 2008.
6
Koniar, D.; Hargaš, L.; Štofan, S. Segmentation of motion regions for biomechanical systems.
Procedia Eng.2012, 48, 304–311. [CrossRef].
7
Uijlings, J.R.; Van De Sande, K.E.; Gevers, T.; Smeulders, A.W. Selective search for object
recognition. Int. J.Comput. Vis. 2013, 104, 154–171. [CrossRef].

74
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

studied area are subject to variable changes such as illumination,


shadows, and other factors.

 Calculation Speed This approach is meant to speed up


the computation performance because the phase of sub-region
searching is merely a preparation for the object recognition step.

Figure 2: Selective Searching

 Background Subtraction One of the most basic techniques for


identifying things in an image, particularly in terms of usability, is
background reduction. The backdrop model must be accurately determined
for this method to be effective. The known background components are
then subtracted from the background model before it is compared to the
current image. The foreground items that are not eliminated are almost
certainly recent additions. Typically, any stationary or sporadically moving
aspects of the scene are considered the background. Time-varying
elements may be present throughout the entire scene, such as moving or
static tree leaves. Systems that monitor objects with a static camera
frequently include a module that removes the background to distinguish
stationary objects from moving ones. Keeping the backdrop model intact is
a crucial and challenging step in the background removal process.
Situations where it is challenging to read or detect the background include
uneven and variable lighting of the scene, changing spectral characteristics
of the illumination, and, as a result, different colours of the objects,
overlapping objects, different camera angles, and object variations within
one category.

75
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

Figure 3: Background Subtraction

 Support Vector Machines (SVM) SVMs are appropriate for


assigning items to N groups since they work by projecting data into
multidimensional space. SVM looks for and determines the plane in which
data is divided into groups. For example, if a vector of features has a
dimension of 2500, SVM will represent it as a point in a space with 2500
dimensions. Talking about SVM working in vectors of features in 2D, Figure
4 shows the pictorial representation of SVM decomposition.

Figure 4: Support Vector Machine (SVM) Principle

 Cascade Classifier (Haar-Like Features) The classification of


stable objects is the main purpose of this classifier. One may discuss the
human body's figure or the face, for instance. The proportions of the
human body, including the hands, legs, head, and other body parts, are
generally the same as those of the face. The shape analysis of the drone
demonstrates that it does not belong in this group of objects. For two main
purposes, a Haar classifier is used to recognise objects8. One benefit is
that, unlike when using insufficient training data, Haar-like features can
accurately describe a region of interest. In comparison to raw pixels, Haar
8
Chen, Q.; Georganas, N.D.; Petriu, E.M. Real-time vision-based hand gesture recognition using
haar-likefeatures. In Proceedings of the 2007 IEEE Instrumentation & Measurement Technology
Conference IMTC2007,Warsaw, Poland, 1–3 May 2007.
76
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

features might, depending on their characteristics, increase or decrease


the variability of data that does or does not belong to a similar class.

The value of the light-to-dark ratio in a scene can be recognised


and defined with accuracy using Haar characteristics. They excel at
common computer vision problems including diverse scenes and changing
lighting. The speed of this classification method is the second justification
for using it, as handling Haar characteristics is thought to be quite effective.

Figure 5: Cascade Classifier

 Machine Learning and Neural Networks It is a complicated


situation when neural networks are used for object recognition. It requires a
sizable set of data to represent the object that has to be recognised in
order to employ a neural network for this task. Some neural network
techniques also require data samples in which the target object is absent.
The weights of the neurons are changed in response to an error that
happens when neural networks classify an object, which causes the
network's total error to decrease over time. This method of training is also
known as the error propagation algorithm. When the network reaches the
selected error threshold, network training is terminated. Another choice is
to halt the training after a specific number of iterations. However, this
method does nothing to address the general inaccuracy of the training.

77
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

Figure 6: Neural Networks

 Tensor Flow (TF) Open-source machine learning platform TF is


used in a wide range of applications 9 & 10. A tensor is a multidimensional
generalisation of vectors and matrices. By using n-dimensional coordinates
of basic data types, TF represents tensors11. Google released this artificial
intelligence-based solution in 2015 (for free use). The dataflow graph
technique is used by TF to depict the calculations. Calculating units are
represented by graph nodes. The edges of the graph, which are
communicated between nodes as tensors (multidimensional arrays)9 & 10,
show the amount of data used or generated by the algorithm12. The
architecture of this system makes it possible to deploy computation on a
range of platforms, including multicore CPUs, graphics cards, and Tensor
Processing Units, quickly and affordably (a computer system also
developed by Google designed primarily for machine learning). A neural
network must be trained using large number of calculations, in order to
classify and recognise objects9. With the help of TF's features, users may
efficiently execute computationally taxing activities like categorisation.

9
Abadi, M.; Barham, P.; Chen, J.; Chen, Z.; Davis, A.; Dean, J.; Devin, M.; Ghemawat, S.; Irving, G.;
Isard, M.;et al. Tensorflow: Asystem for large-scale machine learning. In Proceedings of the 12th
{USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 16), Savannah,
GA, USA, 2–4 November 2016.
10
Abadi, M.; Agarwal, A.; Barham, P.; Brevdo, E.; Chen, Z.; Citro, C.; Corrado, G.S.; Davis, A.; Dean,
J.;Devin, M.; et al. Tensorflow: Large-scale machine learning on heterogeneous distributed systems.
arXiv2016, arXiv:1603.04467.
11
Creative Commons (CC) License. Tensors. Available online: https://www.tensorflow.org/
programmers_ guide/tensors (accessed on 28 April 2018).
12
Creative Commons (CC) License. Graphs and Sessions. Available online: https://www.tensorflow.
org/ programmers _guide/graphs (accessed on 28 April 2018).
78
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

Machine-Learning Approach

 Finding a neural network that is effective at recognizing multi-rotor


drones is the goal of using machine learning techniques. We have selected
the TF machine learning platform for this since it is both readily available
and cutting-edge. The aim is to select a suitable type of network, overtrain
it using adequate network parameters, and then assess its success rate in
classifying objects moving in the sky into the required classifications. The
phases of the suggested solution are listed below:-

 Preparing data for training and evaluation.

 Selection of detection model.

 Files for training/Big Data Analytics.

 Training.

 Export of the trained model to a frozen graph format.

 Creating an application to test the detector.

 Preparing Data for Training and Evaluation A sizable data set is


needed for neural network training to be successful. Photographs of the
subject should be taken in various lighting, spatial, and other conditions. It's
also crucial that the data shows every form of drone that has been defined
in different contexts, rotations, and circumstances. The amount of data
gathered ought to be enough to train the network and identify the target
item. Annotations specifying the precise location of the object in the scene
are necessary in addition to the data collected.

 Selection of Detection Model Since building new identification


model would require a lot of time and resources to get the desired results,
detection models trained on the "Common Objects in Context dataset" are
to be used after evaluating the alternatives. Approximately 2,000 annotated
photographs of various classes (such as aeroplanes, birds, kites, and so
on) that can be used in instruction should be included in this collection. On
the other hand, the drone is not one of them. Since all of the training
parameters would be suited to distinct detection, developing our own
detection model would probably be more reliable than using a model that
has already been trained. These models each have a configuration file that
serves as the model's training data source. It is crucial to keep the
application in mind when selecting the model's input parameters. Models
like Faster Regions with Convolutional Neural Networks (R-CNN) use the
selective search strategy to find likely objects in a scene when the detector
has to be more accurate. These models, however, give less importance to
79
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

processing time. This model will be used in this instance of automated


drone detection. On the other hand, Single Shot MultiBox Detector (SSD)
models give processing time priority.

 Files for Training/Big Data Analytics There must be enough


data for the model to be trained correctly. About 2000 drone images will
therefore be acquired from freely available online resources. The next step
is to create an annotation for each photograph that features the drone,
which serves as the object of interest. 80 percent of the data should be
used for model training, and the remaining 20 percent should be used for
model testing. The publicly available labelIing tool can be used to annotate
and creates a file called an.xml. This file is used to extract the information
that was used to create the Tensor Flow Record (TF Record) file. The TF
training library uses a simple binary file for this purpose. It is thereby
housed in a single memory block, enabling quicker data access. A.csv file
is used to save the data. This file displays the name of the file where the
drone object is placed, as well as the object's height, width, class name,
and coordinates.

 Training The training script was provided by the library's creators


and can be found in the official Github source. An average of one second
was needed to complete one training step (using TF in Google Colab Pro
Plus with GPU support). After a total of 5,000 steps, a model with enough
reliability and viability can be produced (about six training days).

Conclusion

 The TF library was used to create a reliable multi-rotor drones


detection system. The Common Objects in Context dataset was utilised for
this. In ideal conditions, a successful detection rate of 99.3% was attained.
The proposed detection principle was found to work well in the tested
scenarios. The trained detection system performed well in almost every
scenario that it was tested in. An issue did not appear until there were
objects with features shared with objects of another class. The detection
model's statistical analysis revealed that the detection rate in this instance
was only 61.7 percent. Since birds should also be unnoticeable, a different
detection model was used. This model was able to discriminate between
drones and birds, two different types of drones. The tests showed that the
detector was much more successful at detecting items if it was trained on
many things.

 This leads us to the conclusion that the presence of items with


identical characteristics is the biggest obstacle to drone detection. A bird,
an aeroplane, a parachute, or a paragliding wing are examples of
something like drones. These things are less common, therefore, a large
dataset could be used to train the detection model. This kind of detector
80
PINNACLE 2022
Artificial Intelligence in Modern Warfare: Drone Detection Using Machine Learning

may be even more efficient than one that has taken two classes of training.
On the other hand, a detection rate of 97 percent of detectors trained in
two classes is adequate for basic security applications. The system's
reliability and success can be compared to those of a regular human
observer. Based on the provided data and settings, this paper suggests a
practical approach for multi-rotor drone detection under ideal flight
conditions. Future research using this methodology might focus on other
instances where the detector could act as both a common observer and a
complex detection system that can be applied in challenging settings.

81
PINNACLE 2022
COMBAT “AI-EFFECTIVENESS”: ROLE OF ARTIFICIAL
INTELLIGENCE IN COMBAT EFFECTIVENESS
- Colonel G Praveen, SM

“Extensive opportunities in Artificial Intelligence (AI) and Data Analytics


calls for the Indian Armed Forces to prepare a perspective plan to embrace these
technologies and place greater thrust on training its personnel to learn these
technologies and also reskill service and civilian personnel with advanced training
and learning management systems”.
- General MM Naravane

Introduction

The battlefields of the future will be dynamic, chaotic, unpredictable, and


uncertain. In such interlinked battlefields in multiple theatres and domains,
commanders would be sifting through a multitude of data and striking a balance
between their leaders and subordinates for the timely provision of information,
decisions, and feedback. In such a time-constrained operational environment
obscured by the Clausewitzian fog of war, effective and unbiased decision-making
by a leader will be a battle-winning factor. Capabilities offered by the latest
technology will aid decision makers and troops in various domains.

The famous Observe, Orient, Decide, and Act (OODA) loop will have
technology-assisted inputs at each stage of Observe, Orient, Decide, and Act.
Faster processing capabilities of Big Data and inferential analytics and seamless
integration of inputs by machines would simplify the complexities associated with
the five Vs of Big Data, ie, Volume, Velocity, Veracity, Value, and Variety. A
complex dynamic of man-machine interaction is the future, which would mandate
increased trust in a machine in a similar manner to that between two soldiers in
the same team. Human machine teams and xAI, or explainable Artificial
Intelligence (AI), are working towards improving this "Trust" factor, which would
impact Combat Effectiveness.

Some Definitions

AI is a branch of computer science that brings together multiple disciplines


with the aim of creating smart machines-devices and systems capable of
performing complex tasks that often require human intelligence, but in a manner
that equals or exceeds the capabilities of humans. Essentially, the aim is to make
intelligent machines that can replicate human behaviour and intelligence in every
sense. Howard Gardner proposed eight types of intelligence and a ninth one
called existential intelligence; a machine which is equipped with at least one of
them can be called intelligent. As we are aware, amongst the three types of AI,
present capability is in the first level of Weak AI, where a machine or a system like
Alexa or Siri can perform a series of predetermined and defined activities
repeatedly. In day-to-day life, the interaction is with one of the sets or subsets
PINNACLE 2022 82
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

associated with AI, like Big Data, Machine Learning, etc, even though all are
loosely termed as AI.

The essential effect of AI will primarily be on the Combat Effectiveness of


the Armed Forces. Combat Effectiveness is defined as the readiness of a military
unit to engage in combat based on behavioural, operational, and leadership
considerations1. It is a combination of tangible and intangible factors like force
(numerical), equipment availability, morale, leadership, training standards, etc. In
other words, combat effectiveness is the capacity of a military force to succeed in
its mission, or task. Since the components of combat effectiveness are quite a few,
the paper will focus on two major aspects; Decision Making and Training.

Figure 1: Specturm of AI

Developments in AI

Countries like the United States of America (USA), China, and Russia have
made rapid progress in the realm of AI in various domains, including that of
military capability. India has undertaken a series of steps to develop and exploit
the capabilities of AI, which include setting up a national level structure and
Centres of Excellence at various institutes. The Defence AI Council and the
Defence AI Project Agency have been established, and efforts are being made to

1
Encyclopedia Britannica at https://www.britannica.com/topic/combat-effectiveness.
83
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

identify and develop projects across multiple domains. Ten Centres of Excellence
and two dedicated laboratories of the Defence Research Development
Organisation (DRDO) are focused on identifying and developing key technologies
and applications. The Indian Army has set up the AI Centre of Excellence at
Military College of Telecommunication Engineering (MCTE), Mhow and added
AI-oriented curriculum at various training institutes.

Even as new technologies are adopted, ethical and moral concerns,


including concerns about privacy, need to be duly factored in. The subsets of
Machine Learning and Deep Learning require huge datasets for establishing a
referential framework. In the present setup, with levels of confidentiality and limited
sharing and recording of events, having a viable training dataset and an algorithm
that can cater for multiple contingencies becomes a challenge. The famous "Death
Algorithm" concept, wherein an algorithm for a self-driving car should save the
occupants or a pedestrian in case of the likely occurrence of a crash, gives one
word datasets insights into the challenges in devising AI systems.

AI can be used to exponentially improve performance in the fields of


Training, Surveillance, Intelligence Collation, Logistics (including Supply Chain
Management), Cyber Security, Arms & Ammunition, etc. A Manekshaw Paper on
"Leveraging AI for the Military" mentions the domains of situational awareness,
lethality, HR, training, survivability, cyber, Information and Electronic Warfare (EW)
and mobility in which AI can focus on.2 The present thoughts are, however, silent
on three important aspects which must be kept track of while incorporating the
various facets of AI; Bias, Trust and Interoperability.

Decision Making and AI Assisted Debiasing

Cognitive biases in human decision-making and judgement are well


documented. Daniel Kahneman spoke of the System 1 and 2 types of brain,
wherein the fast and intuitive System 1 attempts to hijack the slow and analytical
System 2 brain, thereby leading to an increased propensity for mental shortcuts or
heuristics and thereby biases. Though there are ways and means for removing
and mitigating biases (debiasing), the awareness of a bias does not always result
in the removal of bias from the decision loop. What is more damaging is the lack of
knowledge of implicit biases that normally occur when the brain adopts one of
those shortcuts (heuristics) in a limited time window under conditions of
uncertainty and risk. In addition to having policies and processes at various levels
in the organisation for systematic debiasing and having feedback mechanisms,
there is a need to adopt "Data Driven Decision Making." This would entail common
data sharing platforms and standard algorithms that can decipher inputs and
analyse both structured and unstructured data to offer insights to a leader or

2
Jadhav, Ajinkya, Leveraging AI in the Indian Army, Manekshaw Paper, 10 June 2021, available at
https://www.claws.in/publication/leveraging-artificial-intelligence-in-the-indian-army/ accessed on
15 February 22.
84
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

decision-maker at various stages of the decision-making process. Biases can be


mitigated in the "Observe" phase of the OODA loop by having machines
synthesise large volumes and varieties of inputs to offer a collated intelligence
picture. In the "Orient" phase, when a decision-maker is going through the "sense-
making" paradigm of the decision loop, a collated and synthesised set of analytic
inputs will help in framing and orienting in a faster timeframe. The "Decide" phase
duly supported by a machine learning based Decision Support System can
supplement the experience and intuition of a leader to arrive at a better decision in
a faster manner, which is essential in a futuristic multi-domain battlefield with no
clear boundaries in the cognitive and physical domains.

On the other hand, increased dependence on AI-enabled systems could


further amplify inbuilt biases in the algorithm or existential biases in the training
datasets. Even though there are checks and balances to systematically review
algorithms for bias and to prevent implicit biases from occurring due to training on
incomplete datasets, the probability and likelihood of human cognitive biases
getting incorporated into the design stage cannot be ruled out. Though some
argue that a biased algorithm will still suggest a more balanced "Course of Action"
than a biased human decision maker, the aspect of "debiasing" deserves far
greater attention than what is being given today. It can be said that both humans
and machines can enter into a "mutually beneficial debiasing strategy" to
improve the quality of decision-making.

Human-Machine Teaming and Trust

In any fighting unit or subunit or what could be called a team, each team
member, over a period, having trained together, is aware of the strengths and
weaknesses of the other. Camaraderie and esprit de corps amongst the team
members finally make the team a combat-effective, cohesive unit. Much more than
anything, a key factor amongst the team members is "trust," which gets validated
when the team has to operate under enemy fire, wherein the life of one is
dependent on the covering or supporting fire from the other member or mate. The
introduction of a machine, a faceless, emotionless member into such a cohesive
unit, impinges on the trust factor. Human beings traditionally resort to
anthropomorphism, which is a habit of ascribing human emotions to other beings,
which is commonly seen at homes with pets. A machine that would exhibit human-
like features, habits, and emotions has greater chances of winning over trust in a
team.

Another trust-winning factor could be transparency in actions. As compared


to a teammate and his or her known likely reactions and outputs under various
conditions, the output from a machine is more mechanical and devoid of any
explanation as to why that output was suggested. XAI, or explainable AI3 is one

3
Explainable AI (xAI) techniques, utilizing abstractions or explanations that provide the user insight
into the AI’s rationale, strengths and weaknesses, and expected behaviour, can supply the human
85
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

field wherein the actions and deliberations undertaken by a machine will be more
transparent to the other teammates, which in the long run is expected to increase
the trust factor, which is an important element for battle winning and thereby being
combat effective. In future warfare scenarios wherein, enemies will target the
cognitive domain as part of the Information Warfare spectrum, this factor will
become a key ingredient.

Jointness and Interoperability

For a Battalion or a Regiment or a Ship to be truly combat effective, it must


have a minimum basic awareness of the others will to fight or operate. At a point
wherein the Services are debating the implementation of Theatre Commands and
exploring more joint avenues to operate together, the AI-assisted technologies of
Augmented and Virtual Reality offer the participants a more integrated chance to
train together. The scenarios and settings can be dynamically changed as per
evolving strategies and weapon profiles of the adversaries, while own processes
can be systematically addressed through such joint simulations and AI-assisted
wargames.

The diminishing number of field firing ranges, ever expanding urban


landscape along the periphery of traditional training areas, cost-cutting measures,
etc can be succinctly overcome by adopting the AI assisted Joint Wargaming and
Simulation. This methodology will bring in much needed awareness, dynamism,
and a purple mindset at various levels of leadership. AI-enabled systems can also
assist in identifying key personnel to operate specific platforms right from
recruitment and training stage which will at some stage obviate the biases or
prejudices that are found in identifying the right man/woman for the job.

AI-Effectiveness in a Battalion/Regiment

Researchers usually consider a set of tangible and intangible factors to


measure combat effectiveness: army size, physical and mental prowess of troops,
leadership, the quality of arsenal and tactical aptness in using weaponry, logistics,
intelligence, medical care, knowledge of topography, etc. Measuring the intangible
factors with high accuracy is difficult, and it is the synergy of many contributing
factors that determines combat effectiveness, and researchers cannot eliminate
perceptual, perhaps non-measurable elements4. Despite having details of
weapons systems, it is impossible to use reliable quantitative criteria to evaluate
the actual skills of military personnel in using those systems, or the cumulative

teammate a representation of the collaborative robot’s behaviour policy and may assist in the human
teammate’s ability to predict and develop a collaboration plan. Available at https://
proceedings.neurips.cc/paper/2021/file/05d74c48b5b30514d8e9bd60320fc8f6-Paper.pdf accessed
on 15 March 2022.
4
Value and Methods of Measuring Combat Effectiveness: A New Approach, available at
https://assets. researchsquare.com/files/rs-408649/v1_covered.pdf?c=1631861993
86
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

effect of weaponry efficiency and tactical efficiency. The Lanchester model5 used
for combat modelling, only considers numerical changes in forces deployed, but
the troop strength must be complemented with skills and morale to estimate the
combat effectiveness of an army. In a traditional attack and defence model, certain
research also focuses on the capability of a unit to hold or capture ground as a
measure of Combat Effectiveness.

Existing System Many factors, including camaraderie and cohesion, play


a major role in the Combat Effectiveness of a unit. Presently, a yearly
Administrative Inspection certifies a unit as "Fit for War" after checking a series of
parameters of performance, including firing, battle preparedness, running
standards with loads and level of subunit leadership, to name a few. During actual
operations, which often occur in some counterinsurgency settings, Combat
Effectiveness of a unit is taken as the capability of that unit to achieve success in
any operation, which effectively has many sub domains, including area
domination, intelligence awareness and collation, zero fatalities, etc. Depending on
the terrain and place of operations, the quantification of combat effectiveness also
varies. Though the intangible factors can be quantified to a large extent using
various models, aspects like skill, morale and leadership remain quantification
challenges. Even standards of training, which can be measured at laid down
levels, fail to comprehensively quantify or capture the Combat Effectiveness of that
unit.

AI at Battalion/Regiment Level At various levels and stages, AI and its


associated tools and capabilities can be introduced at various levels and stages to
both enhance and measure skills and standards, while simultaneously being
amenable to further modifications as per dynamic demands of an ever-evolving
battlefield and operational challenges. Mental and physical immersion though
Virtual Reality (VR) and Augmented Reality (AR) combined with haptic
technology66 can add realism and dynamism to training, while simulations and
wargaming models can reduce costs, improve adaptability, and improve inter-
operability. VR can be used in the domains of education, medical, and combat
training. The COVID pandemic saw large scale increase in online teaching
platforms, with many educational institutions using the available technology to
make learning more immersive. Complex concepts could be explained using 3D
online models, while VR systems could help soldiers achieve multiple levels of
education and skill development. Healthcare aspects including combat medical
care can be better learnt and practised by all soldiers. Using the capabilities of
Natural Language Processing, conversational AI systems can be set up that can
engage a soldier on various issues ranging from basic documentation, aspects of

5
https://www.maa.org/press/periodicals/loci/joma/an-interactive-use-of-the-lanchester-combat-model-
the-lanchester-combat-model.
6
https://www.smithsonianmag.com/innovation/heres-what-future-haptic-technology-looks-or-rather-
feels-180971097/.
87
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

field craft and tactics, basic medical clarifications, tenets of Professional Military
Education (PME), etc.

Anthropomorphising, or the human tendency to attribute human feelings


and emotions to other beings, is both a boon and a bane. While it helps in
developing an intimate bond with an otherwise emotionless machine, it could have
a psychological impact similar to how it was described in the movie "Her", where a
lonely human falls in love with a machine which is replicating human emotions
based on backend neural networks and reinforced learning. ELIZA, a
psychotherapist chatbot developed by Weizenbaum in the 1960s, replicated
responses based on key words typed in by the patient but had the capability to
reduce stress levels as the human beings felt they were conversing with someone
who paid attention to their thoughts. Though the capabilities on offer by such
systems which can replicate human responses are tremendous in terms of trust
building and stress busting, the psychological aspects need to be factored in.

Capability Spectrum Many articles have already identified and


suggested the domains in which AI can make an impact in the Armed Forces.
Situational awareness, lethality, Human Resource (HR) management, training,
survivability, cyber, information and EW and mobility are some of them. What
needs to be simultaneously developed is in the psychological domain by agencies
like the Defence Institute of Psychological Research (DIPR) to analyse the impact
and evolve mitigation measures regarding the impact of AI systems on a soldier’s
psyche.

The Future

The moot point is about what will actually change at Battalion/Regiment


level. Will it change the way we work and operate presently? There will indeed be
paradigm shifts in the way we work and operate in terms of efficiency and
transparency. There will also be concomitant demand for change management
and technological skill enhancement. There is a need to set up Human-Machine
Interface Teams (HMiT) at each level, which will facilitate this transition. Starting
from Services Selection Board and Recruitment Centres, AI-assisted software can
help in identifying individuals with the right aptitude and uptake to work in such
human-machine interface teams. These teams can then work in synergy to identify
actionable sections or domains at each level and then collaborate with various
Centres of Excellence to produce need-based machines or systems.

The capabilities of AI and the roles of all associated systems of AI need to


be introduced as part of PME at tri-services level. An integrated multi-domain
fighting force, which is the demand of future battlefields, cannot have commanders
who are unaware of the capabilities that these emerging technologies can offer.
PME needs to address these requirements at the level of officers, JCOs, and
troops. Even as agencies like DIPR devise psychological measures to embrace
machines and systems, there is a need to revamp the present training
88
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

methodology at unit/subunit level, which focuses on Mission Essential Tasks


(MET) and Mission Essential Task Lists (METLs). The HMiT can identify various
stages where technology can enhance and augment success in operations or for
various tasks. After all, success in a task is one of the prime determinants of
Combat Effectiveness.

Even as AI in its present form transforms from "Weak AI" to a stronger


variant with deep learning, neural networks, etc, concepts like "artificial intuition"
and "quantum cognition" will gain more traction. A future-ready force preparing for
multi-planar, multi-domain operations has to be Combat AI-effective and this can
become a reality through systematic, sustainable, incremental yet parallel, nudges
at various levels of the organization. Trust amongst the members of a team is
essential, though an intangible factor like morale needs to transcend to human-
machine trust with adequate safeguards against systemic and implicit biases.

Bibliography

1. Thomas B Allen, War Games: The Secret World of the Creators, Players, and Policy
Makers Rehearsing World War III Today: New York, McGraw Hill, 1987.

2. Kritika Roy, Advances in ICT and the Likely Nature of Warfare: KW Publishers, 2019.

3. Leonard Mlodinow, Subliminal: Penguin, 2012.

4. Daniel D Wheeler, Irving L Janis, A Practical Guide for Making Decisions: The Free Press,
1980.

5. Ed. GJ David Jr, TR McKeldn III, Ideas as Weapons: Pentagon Press, 2009.

6. Mehrabi, Ninareh and Morstatter, Fred and Saxena, Nripsuta and Lerman, Kristina and
Galstyan, Aram, A Survey on Bias and Fairness in Machine Learning, 2021 available at
https://dl.acm.org/doi/abs/ 10.1145/3457607 accessed on 14 Feb 2022.

7. Implementing a Software Modeling - Simulation in Military Training, available at


https://www.proquest.com/scholarly-journals/implementing-software-modeling-
simulation/docview/881226834/se-2?accountid=139958.

8. Tomalin, M., Byrne, B., Concannon, S. et al. The practical ethics of bias reduction in
machine translation: why domain adaptation is better than data debiasing. Ethics Inf Technol 23,
419–433 (2021). https://doi.org/10.1007/s10676-021-09583-1.

9. AI Is Biased. Here's How Scientists Are Trying to Fix It, 19 Dec 2019 available at
https://www.wired. com/story/ai-biased-how-scientists-trying-fix/.

10. Artificial intelligence policy in India: a framework for engaging the limits of data-driven
decision-making, available at https://royalsocietypublishing.org /doi/ 10.1098/rsta.2018.0087
accessed on 20 Mar 2022.

11. National Strategy on AI available at https://indiaai.gov.in/documents/pdf/National Strategy-


for-AI-Discussion-Paper.pdf accessed on 14 Mar 2022.

89
PINNACLE 2022
Combat “AI-Effectiveness”: Role of Artificial Intelligence in Combat Effectiveness

12. Rohan Paleja, Muyleng Ghuy, Nadun R. Arachchige, Reed Jensen, Matthew Gombolay,
The Utility of Explainable AI in Ad Hoc Human-Machine Teaming, available at
https://proceedings.neurips.cc/paper/2021 /file/05d74c48b5b30514d8e9bd60320fc8f6-Paper.pdf.

13. Williams, M. B. S., 2010. Heuristics and Biases in Military Decision Making. Military Review,
5(September-October), pp. 40-52.

14. Saini, C. S., 2008. Role of Intuition in Military Command. Journal of Defence Studies,
1(Winter), pp. 75-87.

15. Jr, M. P. T., 2014. Shaping and Adapting : Unlocking the Power of Colonel John Boyd's
OODA Loop. [Online] Available at: http://www.pogoarchives.org/straus/shaping-and-adapting-boyd-
20150422.pdf[Accessed
15 July 2021].

16. Jung, C., 1971. Psychological Types. First ed. New Jersey: Princeton University Press.

17. Kahneman D, S. P. T. A., 1982. Judgement under uncertainty : Heuristics and Biases. First
ed. Cambridge: Cambridge University Press.

18. Kahneman, D., 2011. Thinking Fast and Slow. First ed. New York: Farrar,Straus and
Giroux.

19. Kahneman, D., 2012. Thinking Fast, Thinking Slow [Interview] 2012.

90
PINNACLE 2022
LEVERAGING ARTIFICIAL INTELLIGENCE (AI) FOR
FURTHERANCE OF MILITARY OPERATIONS
IN THE INDIAN ARMY
- Brigadier S Balakrishnan and Lieutenant Colonel Bikash Biswakarma

“Computers will overtake humans with AI within the next 100 years. When
that happens, we need to make sure the computers have goals aligned with ours”.
- Stephen Hawking

Introduction

Artificial intelligence (AI), a niche technology is rapidly growing and is


capturing the attention of commercial investors, defence intellectuals,
policymakers, and international competitors as evident by its wide application in
virtual assistance in the form of "Alexa" to Information Warfare (IW) in the form of
‘Deepfake’ in recent Russia-Ukraine conflict. Even in the Approach Document
published by NITI Aayog on "Responsible AI for all" in February 2021, it has
been brought out that the National Strategy for Artificial Intelligence (NSAI) has
successfully brought AI to the centre-stage of the reform agenda of the
Government by underlining its potential to improve outcomes in sectors such as
healthcare, agriculture, and education. Furthermore, the NSAI underlines the need
for a robust ecosystem that facilitates cutting edge research to not only solve
societal problems and serve as the test bed of AI innovations, but at the same time
enables India to take a strategic global leadership by scaling these solutions
globally. AI is acknowledged to be one of the foremost dramatic technological
game-changers of our age. It is also referred to as Industrial Revolution 4.0, and
the Government of India has formulated a Task Force to prepare our country for
the upcoming revolution. Defence and security will not be an exception.

Leveraging AI

Leveraging upon AI could fundamentally change the character of warfare


and cause a shift in focus from ‘information’ to ‘intelligence’. Assuming that IA
can leverage the opportunity by the year 2025, then the time to incubate and
induct AI is now, as the gestation period for project development in such niche
technology is phenomenal. Therefore, a review of certain pre-requisites, enablers,
and implementation challenges in the context of AI would be
appropriate. Elaboration follows as under:-

 Pre-requisites The pre-requisites include infrastructure, datasets,


organisational structure, matching financial support and enabling policy
framework.

PINNACLE 2022 91
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

 Enablers of AI Enablers of AI include inter alia Government


initiatives to promote AI, a robust and participative data sharing culture, a
deep pool of national talent in science and technology, scope for Public-
Private-Partnership (PPP) and an indigenous R&D/industrial base.
Strategic partnerships for transfer of technology/knowledge coupled with
spread of digital awareness will only act as catalysts.

 AI-Implementation Challenges A number of challenges besiege


the early implementation of AI in the Indian Army (IA). These include the
promulgation of an enabling policy framework; organisation and HR issues;
technology base and infrastructure to include datasets; and the need for
synergy amongst stakeholders. AI applications will require sustained R&D,
handholding by the industry and academia, and will involve substantial
gestation periods for tangible yields. Also, like any other exciting but
nascent technology, AI too will course through inherent uncertainties and
the associated risk of failure, which needs to be accepted.

Implementation Strategy

 Army Headquarters Computer Centre (AHCC) to take on large


projects with extensive R&D, field trials and finally implement AI solutions
pan Army. The organisation will be responsible to define problem
statement, working out scope and reach of projects and carry out R&D.

 HR Management It is necessary to empower the proposed


organisation created under the Directorate General of Signals for AI
implementation with quality manpower. Talented individuals can be
identified and trained with advance courses to enhance their
capabilities. Post-graduation in AI has already started and officers
are being trained by premier institutes like the Indian Institute of
Technology (IIT). It is necessary to create a pool of talented experts
to work on specific projects with timelines. The provision of hiring
and employing a few talented individuals from outside the IA should
also be enabled.

 Creation of Infrastructure Having created an organisation


with talented manpower, it is necessary to develop infrastructure for
undertaking AI Projects. Following issues need to be addressed
while creating top-class infrastructure (some of it is already in the
pipeline):-

 High Powered Computing Lab AI engines and


algorithms require high processing and computing Graphical
Processing Unit (GPU). It will be necessary to create a
facility with number of high powered engines. A central
facility should be created for all R&D and field trials.
92
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

 Data Management Data forms the building block


for any AI application. The dataset must be tru, authentic,
possess high degree of integrity to provide tangible inputs to
the AI enginge. Different applications like MISO, CICG,
FPMIS and similar other tools generate and store enormous
data, which can be used for the development of AI engines.
However, there is a need to harmonise the available data
through a well-defined framework and make data available
with the "right" data flowing out as the Single Source of
Truth (SSOT) at the right time in a secure manner to the
rightful user. Various systems, such as IA Certifying Agency
(IACA) for token-based access control and Identification and
Authentication Management (IAM) for role-based access
control, are also in place or in the pipeline to ensure the
authenticity, confidentiality, and integrity of data. Once
Network for Specturm (NFS) is implemented, all data
management will be handled by the one Central Data
Centre (CDC) and the six Regional Data Centres (RDCs).
Data will reside at the CDC and RDCs in a distributed
manner and will be accessed over the NFS backbone. The
proposed management of data in IA for AI application is
depicted in Figure 1.

Figure 1: AI Application to Ride Over Complete Database

Areas for Project Development in AI

Intelligent Decision support system (IDSS) The DSS inherently


provides technical insight and prowess, primarily through a software-related
interface. It provides military commanders an excellent tool for dealing with initial
93
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

checks of decision making along with comprehensive assistance later in analysing


key concept data filtering and structuring, relieving the human element of the
mundane, routine, and repetitive tasks. When AI techniques are utilised in the
development of alternatives, the resulting system is referred to as an Intelligent
Decision Support System (IDSS). AI attempts to mimic human decision-making in
some capacity. Advances in AI have shown significant promise in assisting and
improving human decision-making, particularly in real-time and complex
environments. With the advancement of AI related technologies, the AI powered
DSS will, to some extent, venture into the autonomous region. Thus, the system
can take independent decisions in a limited way. This has the added advantage of
compressing the OODA loop.

AI tools such as Fuzzy Logic, Case-based Reasoning, Evolutionary


Computing, Artificial Neural Networks, and Intelligent Agents, when combined with
DSS, provide powerful aids in solving difficult applied problems that are often real-
time, involve large amounts of distributed data, and benefit from complex
reasoning. One way of looking at intelligence is that it is primarily concerned with
rational action, so that an intelligent system would take the best possible action in
a situation. IDSS could be effectively implemented in IA at all verticals to assist in
decision making during fog of war.

Command and Control (C2) With the growing complexity of military C2


in the future battlespace, modelling the C2 is recognised as one of the most
challenging areas. AI will play a vital role in optimising the complexity of command
and control models, which involve reasoning, decision-making, planning and other
high-level intelligent behaviour. In recent years, AI has played an increasingly
important role within modelling C2, which includes the following: -

 Representing C2 domain knowledge.

 Providing ‘intelligent’ simulated environments.

 Facilitating the coordination and communication of agents.

 Determining the validity and reliability of model.

 Interpreting simulation results.

C2 modelling is knowledge intensive. The knowledge includes command


rules according to regulation of battle, the higher level of command orders,
information of battlefield, weapons and equipment manuals, domain-related
documents and experiences, and related experimental data etc. Agent-Based
Modelling (ABM) is a recent simulation modelling technique that consists of
modelling a system from the bottom-up. Such a bottom-up approach captures the
interactions taking place between the system’s constituent units. The primary
benefits of using agent models are that emergent behaviour can be produced from
94
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

the interactions of individual entities, that are easy to manipulate and thus can
cover a large portion of the analytical landscape, and that they eliminate the
logistical headaches associated with conducting human-based experiments. AI,
agents include reflex agents, goal-based agents, utility-based agents, and learning
agents, etc. The structure of the command entity is shown in Figure 2 that
captures the key C2 processes.1

Figure 2: Structure of Command Entity

With the development of machines with human-like intelligence, more


advanced AI techniques will be applied in the context of C2. In this way, IA can
take forward research initiatives and, in collaboration with academia and industry,
implement AI models for C2 structure. In the age of hybrid warfare and grey zone
operations, these AI models too could be utilised to assess the capabilities of an
adversary in terms of kinetic, Electronic Warfare (EW), Information Warfare (IW),
and cyber domains to arrive at the most suitable option to commanders. The AI
engine could access its own and adversary capabilities and suggest the most
suitable options during various brainstorming sessions like war games, and the
same algorithms can be further trained in real-world scenarios. Adversary
capability, including intangibles, can be factored to work out an impact affecting
various stages of battle.

Cyberspace Operation AI is likely to be a key technology in advancing


military cyber operations. Conventional cyber security tools look for historical
matches to known malicious code, so hackers only have to modify small portions
of the code to circumvent the defences. On the other hand, AI-enabled tools can
be trained to detect anomalies in broader patterns of network activity, thus
presenting a more comprehensive and dynamic barrier to attack.2 AI-enabled
cyber tools can be developed with AI algorithms which could autonomously detect,
evaluate, and patch software vulnerabilities before they exploit the network, all
within a matter of seconds rather than the usual months. Progressing on similar
lines, Army Headquarters Computer Centre (AHCC) has conceived a project on
Security Information and Event Management (SIEM) wherein the flow of device
logs from all the network devices and endpoints of the Army Data Network (ADN)
95
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

has been integrated into the hierarchical structure from units to higher formations
to Army Headquarters. Based on use cases, anomalies are being detected and
suggestive measures are recommended for ensuring a robust defensive posture of
ADN. Implementation of Security Orchestration and Automatic Response (SOAR)
to streamline security operations of an organisation is one of the suggested
methods for threat and vulnerability management, incident response, and security
operations automation. The AI-enabled cyber tools could provide a distinct
advantage in defending our networks in future cyber operations.

Information and Electronic Warfare

 Information Warfare AI might also play a vital role in the


information war. Many fear that AI techniques such as Deepfakes—audio,
video, text, and images that are created to show something that did not
necessarily happen, or never occurred—will be a game changer in today’s
warfare. Application of Deepfake as a weapon of IW for misinformation has
been observed in the recent Russia-Ukraine war where videos of the
Russian President are circulating on social media with comments which
were never made. This AI tool is a potent weapon which could change the
course of war and has to be diligently used. A digital footprint to include
purchase history, social media like Twitter, Facebook and openly available
information of senior officials and their relatives of adversaries could be
used to create a comprehensive behavioural profile and used for targeted
information operations. Surveillance networks along the borders could be
used to gather information and create a comprehensive database of
adversaries, and AI techniques could be used to derive intelligence, which
could help in IW.

 Electronic Warfare (EW) EW is one of the most important


characteristics of modern battles. EW can affect a military force's use of the
electromagnetic spectrum to detect targets or to provide information.
Recent developments in AI suggest that this emerging technology will have
a deterministic and potentially transformative influence on combat power.
AI-driven algorithms can be very effective in diverse domains of EW, like
processing radar signals for efficient recognition and classification of
emitters, detecting jammers and their characteristics and developing
efficient anti-jamming algorithms. AI techniques can also enable EW
systems to operate autonomously. Appropriate application of AI in EW can
help to mitigate the attempts by adversaries to hinder their communication
networks, including GPS, satellite signals, etc. AI can reduce the cognitive
burden and improve EW effectiveness for multi-domain operations. This
will rank the incoming data quickly and accurately in order of priority to the
war fighter so that less important signals can be removed. It is also useful
in processing large volumes of data, thereby recognizing its patterns and
deriving meaningful information.3 A generic block diagram of an AI-enabled
EW system is shown in Figure 3. Cognitive technology plays an important
96
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

role in the electromagnetic spectrum and has to be leveraged as an


independent AI module development project, not only for the radio
frequency but also throughout the entire spectrum.

Figure 3: Basic Block Diagram of an AI Based EW System

 Algorithmic Targeting Another major use case of particular


value in a tactical context, is the use of AI in developing rapid and accurate
Automatic Target Recognition (ATR) systems: This is useful since visual
identification of targets by human pilots requires flying at close approach,
putting aircraft at risk from anti-aircraft systems; conversely, although radar
in principle enables the identification and engaging of ground targets at a
standoff distance, this comes at the cost of unacceptably high false-alarm
rates or collateral damage. ATR systems can thus, combine the best of
both worlds, providing highly accurate aircraft fire support from a standoff
distance. Key tactical requirements for such a capability are that it can offer
low false-alarm rates in complex environments-ensuring it is not easily
thrown off by decoys or mistakes-and that it can rapidly improve its learning
capabilities on the basis of sparse or limited training data. DARPA‘s Target
Recognition and Adaptation in Contested Environments (TRACE) research
program is already researching on a machine learning based ATR
programme.

 Situational Awareness and Understanding Although drones


have greatly aided force application in asymmetric conflicts, they have a
number of operational limitations including low flying speed and
vulnerability to air defence systems. Increasing the autonomy of unmanned
systems will strengthen their survivability, enable higher-end performance,
and improve their effectiveness in patrolling or monitoring areas. This feeds
into an increased ability for militaries or states to cover far greater areas
with sensors at a greater cost-effectiveness than human troops. The
enhanced situational awareness enabled by more autonomous and
survivable drones can strengthen the security of bases and experts have
97
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

suggested, could potentially lead to greater stability between states by


enhancing monitoring of contested areas, reducing the viability of covert or
hybrid operations.

 Target Systems Analysis (TSA)/Target Audience Analysis


(TAA) TSA and TAA are intelligence-related methods used to develop a
deep understanding of potential areas for operations. They involve the
analysis of reports, documents, newsfeeds, and other forms of
unstructured information. Further, field AI systems could provide
probabilistic forecasts of enemy behaviour, anticipate and flag bottlenecks
or vulnerabilities in supply lines before they occur, and suggest mitigation
strategies; draw on data (eg, weather conditions collected by drones) to
examine factors affecting operations and assess the viability of different
mission approaches. Natural language processing programmes can filter
social media and news to identify strategically salient themes-or,
conversely, can text-mine the mission reports of their own and allied forces
in order to identify common themes or patterns in engagements. Projects
could be developed in NLP which can recognise and translate languages
like Mandarin and associated dialects in real time. This will not only help in
interpretation during high-level meetings with adversaries but also be an
efficient tool for IW.

 Electronic Medical Records Analysis and Optimizing Medevac


Narrow AI agents could aid in the medevac of injured personnel from
combat zones. Such systems can query up-to-date databases and cross-
reference these against live intelligence from other forces in the area.
Combining information on the severity of injuries, the length and security of
available exfiltration routes, landing sites, and weather conditions, and the
projected rate of medical emergencies in the coming days, such a system
can perform a preliminary triage and determine the optimal means of
evacuating casualties, increasing the efficiency and safety of medical
evacuations and saving lives. Also, medical records can be digitised and,
based on the profile posting of an individual, a detailed forecast of the force
level available to operate in High Altitude Area (HAA) could be available at
any instance of time. The data will also help in accessing the requirements
of various types of medicines required to be stocked at different places
both during operations and peacetime.

 Documents Classification and Crypto-Preserving Intelligence


Sharing During and after operations and exercises, defence
organisations have a need to move information between security domains.
This requires the careful checking of the content to ensure no highly
classified data or information will be moved or revealed to a less highly
classified user or system. While aspects can be automated, the checking
often requires a human to conduct the check. This can be very time-
consuming and is prone to inaccuracy. Algorithms may be developed to
98
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

more accurately read, understand, and verify if the content is safe to send
from one domain to another. It also helps to minimise data aggregation
risks and reduce the risk of inadvertent leaks.

 Logistics Management Large quantities of data are required to


be shifted to make decisions regarding supply, transport, communications,
and so on. Using AI and Machine Learning (ML) in one or more areas of
logistics could help speed up that process and make it more agile. 4 Few
areas where AI could be leveraged to optimise logistics management in IA
are as under:-

 Cloud Service A central repository of information might


sound like a bad idea for the military, but in the world of logistics, it
breeds efficiency. It means savings in time, effort, and money if
logisticians have all the information they need to make informed
decisions when moving supplies and equipment to support troops.
The NFS, CDC and RDCs will assist in the same.

 Supply Chain Management AI will have a future utility in


the field of military logistics. Sensors can be placed at critical
locations/shipments and AI based algorithms can be utilised to
analyse supply chain flows for repair parts distribution, attempting to
determine the most time-and cost efficient means to deliver
supplies. This will potentially generate even greater cost savings in
a shorter period of time. The same can be incorporated in existing
applications of IA.

 Driverless Resupply Drones are attracting attention in the


AI space, but mostly for surveillance. We can look for self-driving
vehicles to resupply outposts and bases, as being done by the US
Department of Defence (DoD) under the Autonomous Ground
Resupply program. This would spare human drivers from the
hardships and risks. Robotic mules could also be used at different
stocking places and warehouses to optimise manual labour efforts.
There are efforts underway in the IA in this field as well.

Conclusion

AI technologies, though still in nascent stages, hold great promise for


facilitating military decisions and enhancing the combat potential. The need of the
hour is to catalyse the development of AI-based systems, utilising the best
available expertise with due indulgence of all stake-holders. It becomes imperative
that the Indian Military take note of the developments in this field and get on board
as soon as possible to derive the advantages that technology will offer in future.
Not only the military but it is also the responsibility of the Government to embrace

99
PINNACLE 2022
Leveraging Artificial Intelligence (AI) for Furtherance of Military Operations in the Indian Army

this technology and provide the industry and academia necessary impetus for
utilisation of AI in every possible military domain is essential.

Endnotes

1. J. Liu and X. Li, "Artificial Intelligence in Modeling Command and Control," 2010 Second
International Conference on Computer Modeling and Simulation, 2010, pp. 383-386, doi:
10.1109/ICCMS.2010.449.

2. Scott Rosenberg, “Firewalls Don‟t Stop Hackers, AI Might,” Wired, August 27, 2017,
https://www.wired.com/story/ firewalls-dont-stop-hackers-ai-might/.

3. „Mayhem‟ Declared Preliminary Winner of Historic Cyber Grand Challenge,” August 4, 2016.

4. Article on “Artificial Intelligence for Military Logistics – Current Applications” by Millicent


Abadicio published on https://emerj.com/ai-sector-overviews/artificial-intelligence-military-logistics/ as
accessed on 26 March 2022.

100
PINNACLE 2022
ARTIFICIAL INTELLIGENCE – RELEVANCE IN
AIR DEFENCE BATTLE
- Major Akhand Pratap

I believe there is no deep difference between what can be achieved by a


biological brain and what can be achieved by a computer. It, therefore, follows that
computers can, in theory, emulate human intelligence - and exceed it.
- Stephen Hawking

Introduction

AI is not a recent phenomenon. Computer scientists like Alan Turing,


Marvin Minsky, and John McCarthy contributed significantly to its theoretical and
technological foundation during the previous 70 years. We are already at the
beginning of the exponential age of AI as businesses learn to unlock the value
locked in massive amounts of data thanks to nearly unlimited computer power and
falling prices for data storage. As a result, by gathering and analysing images,
sound, and speech, computer vision and audio processing may actively
comprehend the environment around them.1 The AI system can analyse and
comprehend the information gathered with the help of the natural language
processing and inference engines. A machine learning system can also act
physically or through technologies like inference engines and expert systems.
These human capacities are enhanced by our capacity for lifelong learning and
adaptation. As these skills become more sophisticated, AI systems are being used
in a wider range of organisations to supplement them. But regardless of the kind of
AI being applied, every application starts with a substantial amount of training
data. Statistical regressions, rule-based data analytics algorithms, and early
"expert systems" were once responsible for this type of performance. However, the
proliferation of robust deep neural networks has given AI the capacity to perform
unanticipated tasks, something that a simple programme lacks.2

Evolution of AI

In the history of AI technology, waves of hope have been followed by


setbacks and "AI Winters," or periods of inaction. Previous innovations have never
been able to fully live up to the anticipation they created, and none have been able
to propel the technology into the mainstream. Today's major shift is that we are
seeing a moment of unparalleled technological advancement in a wide range of
industries, leading us to believe that the "AI Spring" has not only begun but will
continue.3

1
Author: Virtul Mittal, https://sharedservicesforum.in/think-big-start-small/.
2
Discussion Paper National Strategy for AI NITI Ayog 2018.
3
Author: Virtul Mittal, https://sharedservicesforum.in/think-big-start-small/.
PINNACLE 2022 101
Artificial Intelligence – Relevance in Air Defence Battle

Exponential increase in digital data is a key factor that accounts for this
optimism. International Data Cooperation (IDC) predicts that by 2025, the amount
of data generated worldwide would increase to 163 (ZB), or a trillion gigabytes, or
ten times the 16.1 ZB generated in 2016.

Figure 1: What is Artificial Intelligence4

Data is to AI what nourishment is to humans, according to Barry Smyth,


professor of computer science at University College Dublin. As the world becomes
increasingly digital, AI advances are continually being fuelled by the exponential
rise of data. Additionally, General Bipin Rawat stated that "the military is the
repository of big data, and there is a need to preserve and institutionalise the
information and conduct predictive analytics utilising AI."

The Chinese government unveiled a roadmap outlining its goal of


dominating AI by 2030 on 20 July 2017. Vladimir Putin openly declared Russia's
intention to pursue AI technologies less than two months later, saying, "Whoever
becomes the leader in this field will govern the world." Similar to this, the January
2018 release of the U.S. National Defence Strategy named artificial intelligence as
one of the critical technologies that will "guarantee the United States to be able to
fight and win the battles of the future."

4
Source: Accenture
102
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Figure 2: Evolution of AI 5

Scope of Use of AI in Military Operations

Almost every imaginable sector of the civil sector has been affected by AI.
It has revolutionised how people live and conduct business, and it is now swiftly
advancing to play a crucial role in contemporary combat. Investment in this area is
the highest in some of the most developed countries when compared to other
sectors.

This investment is mostly used for thorough research and development of


cutting-edge technologies for military purposes, including AI. Military systems with
AI are capable of efficiently handling large amounts of data. Due to their enhanced
computation and decision-making powers, such systems have also increased self-
control, self-regulation, and self-actuation. The market survey and report from
markets, which states that the AI in Military Market is projected to grow from USD
6.26 billion in 2017 to USD 18.86 billion by 2025, at a CAGR of 14.75 percent. In
the upcoming years, the AI in military market is anticipated to be driven primarily
by big data analytics, artificial neural networks, and advanced analytics.

5
Source: Discussion Paper National Strategy for AI NITI Ayog 2018.
103
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Figure 3: Prediction of AI in Military Market

Use of AI in Militaries around World The United States of America


(USA), China, and Russia are some of the front-runners in implementing cutting-
edge technologies to bolster their military capabilities. The United States
Department of Defence published its first AI policy in 2019. It called for boosting
the use of AI systems across the military, from problem prediction to decision-
making, by investing in and partnering with private institutions in AI research.
Project Maven, Defense Advanced Research Projects Agency’s (DARPA) Squad
X Experimentation Program, and the OFFSET programme are some of the most
well-known AI-based programmes that have previously been implemented in the
US military.

The Chinese government has also pushed for more advancements in


military technologies that are based on AI and machine learning. Military-civil
fusion style innovation, as advocated by President Xi Jinping, has also been
included in Chinese National Plan. The "Military-Civil Fusion National Defence
Peak Technologies Laboratory," established by Tsinghua University, is one of the
projects under development. Another initiative is the creation of the Blowfish A2
model in cooperation with Ziyan Unmanned Aerial Vehicle (UAV). The company
claims that its Blowfish A2 model can automatically carry out difficult combat tasks
like fixed-point time detection and precise target strikes. Notably, China published
its State Council AI Plan in 2017, outlining the nation's ambitious initiative to build
an AI sector valued at 150 billion RMB.

104
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Russia is renowned for its military might. The Foundations for Advanced
Research Projects is the nation's version of DARPA. Higher hierarchies have been
seen supporting the development of AI-based technology on the front of AI for the
military. According to reports, in order to conduct more effective Information
Operations, the Russian defence forces are considering leveraging AI, big data,
and machine learning. Currently, Russia makes significant investments in AI for
the commercial and defence sectors in order to detect, analyse, and refute
misinformation.

India is implementing AI-based innovation in its combat and surveillance


initiatives in small increments. The Ministry of Defence created the high-level
Defence AI Council (DAIC) in 2019 with the purpose of giving strategic guidance
for the adoption of AI in defence. Leading the collaboration between the
government and the industry for the application of such technologies is one of
DAIC's main responsibilities.6

Domains of AI in Military

 Training Multidisciplinary domains like training and simulation use


systems and software engineering ideas to build models that let soldiers
practise using different fighting systems that are used in military operations.
Numerous sensor simulation programmes have already been started by
the US Navy and Army. Additionally, simulations for training can be made
that are efficient, realistic, and dynamic using techniques from augmented
and virtual reality. Both real soldiers and virtual agents benefit from
improved combat training thanks to reinforcement approaches.

 Surveillance The extraction of useful intelligence from linked


equipment, such as radars and automatic identification systems, can be
aided by AI with geographical analysis. By alerting the appropriate
authorities, this information can assist in the detection of any unlawful or
questionable activity. Target identification and classification can also be
assisted by AI-powered and IoT-connected robots using computer vision.

 Arms and Ammunition Modern weapons now have AI


technology built in. For instance, advanced missiles can analyse, identify,
discriminate, and prioritise targets without human assistance.

 Cyber Security As a new dimension of warfare, cyberspace is


currently regarded in defence circles as the third warfront. The security of
the entire region could be seriously jeopardised by a corrupted and
malevolent network. Machine learning is being used by defence agencies
to anticipate threats and guard against unauthorised breaches.

6
https://analyticsindiamag.com/what-are-the-scope-and-challenges-of-using-ai-in-military-operations/.

105
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Categorizing the network as normal or intrusive is the typical method for


performing this intrusion detection. AI-based methods aid in improving the
categorization accuracy.

 Logistics The success of a military operation is largely dependent


on a number of crucial aspects, one of which is logistics. The amount of
effort, time, and error is reduced by integrating geospatial analysis and
machine learning with the military's logistics systems.

IMPACT ON AIR DEFENCE BATTLE

AI Enabled Air Threat and Weapons AI is likely to impact future


military aviation becoming part of the next generation hybrid, rotary and fixed wing
unmanned systems. It will challenge the conventional monolithic systems thus
redefining air power strategy. Salient aspects of Air Threat envisaged are listed
below: -

 AI Enabled Fighters A within-visual-range, air-to-air combat


drone that leverages AI for tactical decision making appears to be a simple
engineering problem.7 In fact, the United States Air Force (USAF) intends
to replicate the AI-piloted aircraft versus human-piloted aircraft experiment
from 2020 in 2024, but this time using real tactical aircraft rather than
simulations. An operational, enhanced AI-enabled short-range, dog fight
capable drone might be cheaper, smaller, and lighter than a crewed
aircraft, and it might not even need to be armed to defend against an
approaching enemy air attack.8

 Precession Guided Munitions and Missiles Long-range


precession missiles and munitions are seen as being a significant threat in
addition to conventional air attack because they may travel greater
distances unnoticed and are difficult to neutralise because of their small
size. Once they have reached their intended target, these missiles may
identify and select their target more quickly and precisely, reducing the
time and chance for air defence system to engage.

 SWARM Attack A Swarm of drones capable of communicating


with one another, combined with an AI-based targeting and identification

7
Joseph Trevithick, ―Navy Establishes First Squadron to Operate Its Carrier-Based MQ-25 Stingray
Tanker Drones,‖ The Drive, 1 October 2020, https://www.thedrive.com/; Kyle Mizokami, ―Russia‘s
‗Hunter‘ is Unlike Anything in America‘s Arsenal,‖ Popular Mechanics, 10 August
2020, https://www.popularmechanics.com/
8
Patrick Tucker, ―An AI Just Beat a Human F-16 Pilot in a Dogfight — Again,‖ Defense One, 20
August 2020, https://www.defenseone.com/; and Secretary of Defense Dr. Mark T. Esper, ―Secretary
of Defense Remarks for DoD Artificial Intelligence Symposium and Exposition,‖ US Department of
Defense, 9 September 2020, https://www.defense.gov/
106
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

system, would wreak havoc on adversaries. It has the potential to attack


any target, including bunkers, people, and equipment.

AI based Sensor Fields and Command & Control Model9

The system envisages deploying a significant number of inexpensive


Internet of Things (IoT) sensors on land, sea, air, space, and cyber domains
throughout the combat zone to keep a close watch on adversaries’ activities. The
Integrated Air Defence System (IADS) concept, consists of a network of
surface-based radar stations augmented by Airborne Early Warning and Control
Aircraft (AEW&C) to detect high and low-flying aircraft, already implements this
principle.

 This idea involves using a significant number of AI-enabled small,


low-cost surface and airborne sensors in addition to the current, expensive,
limited-number sensor deployment. The partially processed data is
transferred through the cloud to a fusion centre and subsequently to the
command-and-control system.

 These networked Air Defence grid would involve numerous short


range radar transmitters and passive IOT sensors capable of detecting
emissions in the acoustic, Ultraviolet (UV), Infrared (IR) extending upto
entire spectrum

 These low cost, small sensors may have low individual


performance, however, the integrated outputs of hundreds of these
sensors, would be able to provide a gap free, coherent air picture which
would allow for tracking, identifying and subsequent engagement of enemy
aircrafts.

 Air Defence IoT sensors on surface may be stationary and


persistent, whereas drones with integrated sensors may have endurance
ranging from a few hours to a day. High-altitude balloons, miniature and
pseudo-satellites are just a few of the new IoT platforms that have the
potential to significantly extend this endurance.

9
https://www.airuniversity.af.edu/Wild-Blue-Yonder/Article-Display/Article/2548127/ai-enabled-war-in-
the-air/.

107
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Figure 4: AI Based Sensor Fields and Command and Control System

 Intruding aircraft will have to avoid using transmitting equipment like


radars, data linkages, and communications systems to evade detection as
a significant number of these IoT sensor employ passive detection
technique. However, typical aircraft emissions like noise, heat, and visual
signatures may still make an aircraft vulnerable to detection by a broad IoT
sensor field. Aircraft may perform manoeuvres to reduce these emissions,
particularly those from the forward section of the aircraft when it
approaches a recognised sensors field. A deep sensor field would allow
identification of a penetrating aircraft from flanks and rear section, even if it
has not been picked up during ingress.

 A fusion facility would receive partially processed data from the


extremely vast IoT sensor field enabled by AI and send it to the cloud for
further processing by AI. The Observe-Orient-Decide-Act model can be
used to process these sensor outputs. The IOT sensors, coupled with
processing of data at fusion centres would form part of the first two stages
of Observe and Orient stage. AI would not only create a comprehensive air
picture in near real time but would also forecast adversaries’ courses of
action and movements.

 The next AI layer handling "Decide," would send a prioritised list of


approaching air targets to engage, the best cross-domain assault
strategies to use, relevant timings, and any disengagement considerations
to the human commander for approval. Humans would continue to be
actively involved in this through in-the-loop or on-the-loop control.

108
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Following human approval, the following AI layer would automatically


allocate the chosen weapons to each target, delivering the necessary targeting
data, assuring friendly force disengagement, confirming when the target was
engaged, and even commanding weapon munition resupply. Act, the last stage,
would mostly be a human function.

Solutions with Available Technology A limited number of AI


applications can be implemented right away because they are being developed in
business industries, such as:-

 Automated Air Sentry With help of pattern recognition and


machine learning independent cameras can be installed at difficult and
un-accessible terrain, where sustenance of radars is difficult due to power
requirements and high-altitude constraints. In our context high altitude
areas of Ladakh and Northeast can be covered by such Air Sentries.

 Data Management Systems with AI Processing The air defence


battle requires a large number of reports and returns to be generated for
correspondence to provide near real time data of tactical and logistic
situation to higher headquarters to include number of aircrafts shot down,
requirement of ammunition, casualty report etc. This data is crucial for
success in any battle as it provides commanders with a clear picture of
battlefield and allows them to take informed decisions. This data can be
transmitted, stored, and processed with an AI enabled central server and
can allow for accessing it in a blink of an eye. For example, imagine an
Alexa kind of application in office which can carry out all staff check and
provide accurate data without delay and chances of human error.

 AI Enabled Anti Drone Systems. The vulnerability of static


installations like airfields, refineries etc. has been adequately highlighted in
recent attacks on such installations across the world. These installations
are now guarded with state-of-the-art Anti Drone Air Defence Systems. IOT
sensors on the periphery of such installations integrated over an AI based
algorithm with passive and active anti drone systems would assist in
neutralizing these attacks effectively.

 AI Based Simulators Air Defence missiles are extremely


expensive and therefore it is a costly proposition to make large number of
missiles available for training. Similarly, train radar operators on techniques
of Electronic Counter Measures and Electronic Counter Counter Measures
there is a need to coordinate with fighter aircraft flying. AI based missile
gun and simulators would allows saving huge training cost an allow large
number of gunners and operators to be trained simultaneously. Real time
jamming operations can be generated on radar simulators thereby
improving training standards.

109
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Challenges in the Future The investment of time, money, and expertise


required to use AI-based technology is a major challenge. India is a growing
economy and requires concentrating its resources for development of human
resource index: access to knowledge, a decent standard of living, and long and
healthy life. A big difficulty is determining how much resources we can afford to
offset for development of such technologies for military. There is a need to identify
essential and dual use technologies which can assist in country’s development as
well as security.10

An ethical conundrum is also raised by the deployment of AI in kinetic


domain. Global organisations and experts have boosted this technology,
unintentionally exacerbating international tensions. One of the arguments is that if
an AI system doesn't work as designed, it could have disastrous consequences. In
reality, a number of human and civil rights organisations want a complete ban on
autonomous defence equipment, particularly weapons.

Recommendations Implementing AI based Air Defence System would


require the following: -

 Development of expertise for producing AI based systems and


understanding of AI for operating such systems.

 Creating a group or organisation within services for faster and


better understanding of user problems.

 Framing Policies for supporting AI and to understand and address


the ethical, legal and societal implications of AI.

Conclusion

Air Defence battle requires management of greater amount of data


handling in limited time frame being short, intense and time critical which further
shortens the decision cycle. AI however, shows promise in providing solutions for
data acquisition, fusion and analysis thereby enhancing the situational awareness
of commanders and reaction capabilities of shooters at the tactical level. AI
integrated Air Defence systems would allow real time identification of friend and
foe (IFF) and would reduce fratricide thereby providing enhanced freedom of
operation to friendly aircrafts. As the air threat envelope enhances from low cost
drones to sophisticated 4th & 5th generation aircrafts, ballistic and cruise missiles a
focused approach in Research and Development is required for incorporating AI in
Air Defence Systems.

10
Author: Shraddha Goled, https://analyticsindiamag.com/what-are-the-scope-and-challenges-of-
using-ai-in-military-operations/.
110
PINNACLE 2022
Artificial Intelligence – Relevance in Air Defence Battle

Bibliography
1. Allan, M Din, ―Arms and Artificial Intelligence‖, Oxford University Press, 1987.

2. Bostrom, ―Superintelligence‖.

3. Clayton M Christensen, ―The Innovator‘s Dilemma: When New Technologies Cause Great
Firms to Fail‖.

4. Topychkanov, Impact of AI on Strategic Stability and Nuclear Risk.

5. Zachary, S Davis, ―Artificial Intelligence on The Battle Field – A realistic appraisal of AI, Big
Data & Machine Learning‖.

6. Discussion Paper on National Strategy on AI by NITI Ayog 2018.

7. Artificial Intelligence and Its Impact on the Indian Armed Forces by Maj Gen PK
Chakravorty.

8. Cummins ML, Article on ―Artificial Intelligence and The future of Warfare‖- Defence
Technology (January 2017, Number 128).
9. Gronlund Kirsten, Article on ―State of AI : Artificial Intelligence, the Military and Increasingly
Autonomous Weapons‖, Future of Life.
10. Kulshrestha S, Dr Rear Admiral, Article on ― Indian Armed Forces Approach To
Managing ISR Big Data‖,IndraAstra Oct 2016 Edition.
11. Talwar Surjeet Singh, ‗Disruptive Technologies : Impact on Warfare and Their Future in
st
Conflicts of 21 Century – Centre for Land Warfare Studies (CLAWS)‘, accessed 12 September
2020.

12. Article on ―What are the Scopes and Challenges of using AI in Military Operations‖.https://
analyticsindiamag.com/what-are-the-scope-and-challenges-of-using-ai-in-military-operations/.
13. Article on ―AI in Air Combat Indian Air Force leads the Defence‖. https://www.financial
express.com/defence/artificial-intelligence-in-air-combat-indian-air-force-leads-the-defense-initiative
/2228352/.
14. Article on ―Using Artificial Intelligence in Big Data‖ by Analytics Insightchttps://www.analytics
insight.net/using-artificial-intelligence-in-big-data/, accessed on 07 Oct 20.
15. Article on AI enabled War in the Air https://www.airuniversity.af.edu/Wild-Blue-
Yonder/Article-Display/Article/2548127/ai-enabled-war-in-the-air/.
16. Article on ―Navy Establishes First Squadron to Operate Its Carrier-Based MQ-25 Stingray
Tanker Drones‖ https://analyticsindiamag.com/what-are-the-scope-and-challenges-of-using-ai-in-
military-operations/.
17. Joseph Trevithick, ―Navy Establishes First Squadron to Operate Its Carrier-Based MQ-25
Stingray Tanker Drones‖ The Drive, 1 October 2020.
18. Article on Russias Hunter is unlike Anything in America‘s
Arsenal https://www.thedrive.com/, Kyle Mizokami, ―Russia‘s ‗Hunter‘ is Unlike Anything in America‘s
Arsenal,‖ Popular Mechanics, 10 August 2020, https://www.popularmechanics.com/.
.
19. Patrick Tucker, ―An AI Just Beat a Human F-16 Pilot in a Dogfight - Again,‖ Defense One,
20 August 2020, https://www.defenseone.com/; and Secretary of Defence Dr. Mark T. Esper,
―Secretary of Defense Remarks for DoD Artificial Intelligence Symposium and Exposition,‖ US
Department of Defense, 9 September 2020, https://www.defense.gov/.

111
PINNACLE 2022
ARTIFICIAL INTELLIGENCE AND FUTURE OF WARFARE:
ROADMAP FOR THE INDIAN ARMY
- Lieutenant Colonel Deepak Kumar Tiwari

Introduction

C4I2STAR (Command, Control, Communication, Computers, Information,


Intelligence, Surveillance, Target Acquisition, and Reconnaissance) is the
acronym for systems that form the cutting edge of today‟s warfare. These systems
define the technological superiority of any nation. The network centric approach to
warfare means utilising information age concepts to integrate and undertake a
different range of missions to improve both the efficiency and effectiveness of
operations.

Recently, the integration of computers and revolution in the field of


information has led to concepts such as the Internet of Things and Big Data
Analytics. This pool of data is best used by autonomous machines and programs.
Every nation with significant military might has declared budgets for research and
development on artificial intelligence (AI) and its military applications in recent
years. This genre of technology is likely to revolutionise elements of C4I2STAR.
The command and control (C2 of C4I2STAR) system was always human-in-the-
loop and required input from the other technologies (communication, computers,
information, intelligence, surveillance, target acquisition, and reconnaissance).
Today, every system is on the verge of using autonomous systems to work in
tandem or remove the human-in-the-loop.

Automatic and Automated System

Before understanding Artificial Intelligence (AI), it is important to


understand automatic and automated systems. Automatic systems are simple
machines that do not exhibit much in the way of "decision-making." Automated
systems are more complex and may consider a range of inputs and weigh several
variables before taking action. An automated system is one in which a computer
reasons by a clear if–then–else, rules-based structure and does so
deterministically, meaning that for each input, the system output will always be the
same (except if something fails). An autonomous system is one that reasons
probabilistically given a set of inputs, meaning that it makes guess about the best
possible course of action given sensor data input. Unlike automated systems,
when given the same input, autonomous systems will not necessarily produce the
exact same behaviour every time; rather, such systems will produce a range of
behaviours.

AI is programmed to do something similar. A computer senses the world


around it, and then processes the incoming information through optimization and

PINNACLE 2022 112


Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

verification algorithms, with a choice of action made in a fashion similar to that of


humans. The technology envisages applications in both civil and military domains
alike. AI has shown many applications in the contemporary world. AI-empowered
cars are already under rigorous testing and they are likely to be common on the
road. In 2017, the social humanoid robot Sophia became a citizen of Saudi Arabia
in 2017. Natural language processing by machines is revolutionising mobile
phones. Autonomous weapons can execute military missions on their own,
identifying and engaging targets without any human intervention. AI enables
intelligent machines that can execute functions similar to human abilities like
speech, facial, object, or gesture recognition, learning, problem solving; reasoning;
perception; and response.1

Investment in AI has seen a remarkable rise in the past few years


(Figure 1). However, the spike in funding and focus is not restricted to the civilian
domain. The United States, Russia, China, and other countries are increasing their
investments in artificial intelligence. Russia and China have published their
national strategies on artificial intelligence, and by 2030 they want to become
leaders in AI technology. The Ministry of Commerce and Industry of India has set
up a task force on Artificial Intelligence to kick-start the use of AI for India's
economic transformation. The report was released on 20 March 2018. One focus
area suggested by this report is National Security.

Integrating AI weapon systems into military platforms has a broad range of


applications but numerous concerns, both practical and ethical. Autonomous
weapon platforms have the potential to significantly reduce the manpower required
to perform a myriad of tasks. This is important given that some of these tasks, like
patrolling and mine clearing, are exceptionally dirty, dull, or dangerous.
Autonomous weapon systems can perform the same task for longer durations
more reliably. They lack human limitations such as fatigue, boredom, and injury.
Furthermore, when placed in a combat situation, an autonomous system would
have the ability to rapidly analyse data and react without human limitations,
including panic and injury. Paradoxically, many of the same benefits can be
interpreted as causes of concern.2 While there is a strong consensus that Artificial
Intelligence will be a game-changer and a key factor in development, there is a
concurrent need to arrive at frameworks that will promote its deployment, taking all
risk factors into account.

Applications of AI in Military

 Force Structuring Because of the military's large structure and


the constant revision of concepts in war fighting, a continuous review of
formation structuring, and employability is required. This process is done
1
Sharma Munish,The Global Race for Artificial Intelligence: Weighing Benefits and Risks, IDSA Issue
Brief, 23 February 2018.
2
https://ww2.kqed.org/education/2016/09/14/are-artificially-intelligent-military-systems-worth-the-risk/
accessed on 29 June 18.
113
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

manually, and in countries with large armed forces, this process takes a
long time. Successful implementation of AI in force structuring might lead to
new concepts of operation that could influence how militaries organise
themselves and plan operations.3

 Intelligence, Surveillance and Reconnaissance (ISR) Due to


the availability of large amounts of data in ISR field, AI has found extensive
utility. AI is planned to automate the work of human analysts who currently
spend hours analysing videos and images for valuable information. AI may
reduce the workload of analysts and improve their efficiency by allowing
them to take timely AI-based solutions. The intelligence agencies have
sponsored projects such as image recognition or labelling to predict future
terrorist attacks or civil unrest based on wide-ranging analysis of open-
source information.4

 Logistics IBM Watson is an AI processor that can validate and


audit transportation decisions to optimise modes of transport for logistic
requirements to reduce cost. The movement of troops and material in a
country as vast as India is a complex task. In the Indian scenario, a trivial
decision of troop transport in valleys based on available effort and available
time for movement needs deliberation and analysis. This task, if automated
with AI, will reduce effort at the operational level. The concept is to employ
AI to determine whether air, land, or sea is the more rational mode for a
given shipment.5

 Maintenance AI has also found unique applications in preventive


maintenance, stocking, inventory management etc. Aircraft maintenance
involves conducting repair tasks when an aircraft snags or has a planned
schedule with inherent delays. Predictive aircraft maintenance based on AI
would allow technicians to perform maintenance on individual aircraft on a
need-basis. This kind of application would need real-time data from
sensors and a large data pool to substantiate decisions for the system.
Such sensors could be embedded in the aircraft‟s engines and other on-
board systems and further feed them into a predictive algorithm to
determine when technicians need to accomplish inspections or replace
parts.6

 Cyberspace Conventional cyber-defence tools look for matches to


previous malicious code, so intruders have to modify small portions of that

3
Artificial Intelligence: Autonomous Technology (AT), Lethal Autonomous Weapons Systems (LAWS)
and Peace Time threats By Regina Surber, Scientific Advisor, ICT4Peace Foundation and the Zurich
Hub for Ethics and Technology Pp 5.
4
HoadleyD,Lucas N, Artificial Intelligence and National Security, April 2018, Pp 9.
5
https://www.c4isrnet.com/home/2017/09/07/army-logistics-integrating-new-ai-cloud-capabilities/
accessed on 10 December 18.
6
HoadleyD,Lucas N, Artificial Intelligence and National Security, April 2018, Pp 9.
114
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

code to defeat this defence. AI-based cyber-defence tools can be taught to


distinguish these minor variations from patterns or behaviour in a network
and detect anomalies. Thus, AI-based cyber defence will present a more
wide-ranging barricade to previously vulnerable attack strategies. These
AI-based tools will allow defenders to be protected against new and
creative means of cyber-attack instead of the simple mechanisms of the
past.7

 Command and Control In the immediate future, AI may be used


to fuse data from sensors in all of these domains to create a single source
of information for decision makers, also known as a common operating
picture. The information available to decision makers comes in diverse
formats from multiple platforms, often with redundancies or unresolved
discrepancies. A common operating picture enabled by AI would combine
this information into one display, providing an intuitive picture of friendly
and enemy forces, and automatically resolving variances from input data.
Later, AI may be used to identify communications links cut by an adversary
and find alternative means to distribute information. As the complexity of AI
systems matures, AI algorithms may provide commanders with viable
courses of action based on real-time analysis of the battlespace, which
would enable faster adaptation to unfolding events.8 In the long run,
analysts believe this area of AI development will likely be especially
consequential, with the potential to improve the quality of wartime decision-
making and accelerate the pace of conflict.

 Autonomous Vehicles All the military services are incorporating


AI into various types of autonomous vehicles. AI applications in this field
are similar to commercial self-driving vehicles, which use AI technologies to
perceive the environment, recognize obstacles, fuse sensor data, plan
navigation, and even communicate with other autonomous vehicles.AI-
based cooperative behaviour, or swarming, is a unique subset of
autonomous vehicle development, with concepts ranging from large
formations of low-cost drones designed to overwhelm defensive systems to
small squadrons of RPAs that collaborate to provide electronic attack, fire
support, and localised navigation and communication nets for ground-troop
formations.9

 Lethal Autonomous Weapon Systems (LAWS) LAWS are a


special class of AI systems capable of independently identifying a target
and employing an onboard weapon system to engage and destroy it with
no human interaction. LAWS require a computer vision system and
advanced machine learning algorithms to classify an object as hostile,

7
ibid, Pp 10.
8
HoadleyD,Lucas N, Artificial Intelligence and National Security, April 2018, Pp 10.
9
ibid Pp 11,12.
115
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

make an engagement decision, and guide a weapon to the target. LAWS,


once activated, would, with the help of sensors and computationally
intensive algorithms, identify, search, select, and attack targets without
further human intervention. Whether a human being can still overpower or
veto an autonomous weapon‟s „decision‟ in order for it to be called LAWS is
also being debated. From a military perspective, LAWS have many
advantages over classically automated or remotely controlled systems.
LAWS would not depend on communication links; they could operate at an
increased range for extended periods. Their higher processing speeds
would suit the increasing pace of combat.

Progress of AI in Military Domain

Global investments in artificial intelligence for economic and national


security purposes are increasingly described as an arms race. China published a
national strategy on artificial intelligence in 2017 that said AI represents a “major
strategic opportunity” and proposed a coordinated strategy to “build China‟s first
mover advantage” and lead the world in AI technology. Russia is investing heavily
as well, especially in the military domain. Reports suggest that the Russian military
is designing autonomous vehicles to guard its ballistic missile bases as well as an
autonomous submarine that could carry nuclear weapons. In robotics, Russia is
deploying remotely piloted tanks, such as the Uran-9 and Vehar, on the battlefield.
China and Russia are not the only actors outside the United States interested in
national security applications of AI. The character of AI technology, like robotics,
makes many countries well-positioned to design and deploy it for military
purposes. Commercial incentives for AI developments and the dual-use character
of many AI applications mean that countries with advanced information economies
are poised to be leaders in AI or at least fast followers. In Southeast Asia,
Singapore is on the cutting edge of AI investments (both military and non-military).
Other Southeast Asian nations are making advances in AI research as well. In the
military domain, South Korea has developed the SGR-A1, a semi-autonomous
weapon system designed to protect the demilitarised zone from attack by North
Korea. AI also provides opportunities for capital-rich countries, which creates
incentives to develop the technology. Israel, a classic example of an advanced
economy with more capital than labour, also funds military AI investments that
would predict rocket launches and analyse video footage.10

Status Update of AI in India

AI has a distinctive "dual-use" nature. While its benefits for humanity are
immense, they all come with a multitude of risks from a multitude of factors. The
situation is not about humans vs. machines but about harnessing camaraderie

10
H Michael, Artificial Intelligence, International Competition, and the Balance of Power, Texas
National Security Review: Volume 1, Issue 3 (May 2018).
116
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

between humans and machines to make better decisions. Today, AI has opened
up a geopolitical debate and intense competition among nation-states.11

 Ethical and Legal Concerns from LAWs Autonomous weapons


function with minimal to no human intervention, selecting and engaging
targets by themselves. Fully autonomous weapons do not yet exist, but an
increasing number of countries are developing or deploying near-
autonomous systems. While no country has acknowledged deploying
autonomous offensive weapon systems, even defensive systems will
profoundly change the way nations think about wars and will directly affect
a variety of areas, including trade and the balance of power. A small but
fierce global debate has generated issues related to the morality of these
weapons and their status under international humanitarian law. A
consensus has not yet emerged, but parties such as the United Nations
(UN) special rapporteur on extrajudicial executions and the state of
Pakistan have already put forward arguments calling for a pre-emptive ban
on the development of such weapons. Other parties, such as the US Air
Force, have argued that autonomous weapons have benefits, including
reducing casualties and improving efficiency in defensive capabilities. Only
the United States and the United Kingdom have released official
documents that clearly lay down their respective positions on autonomous
weapons.12 The other countries expressed their positions at the three
gatherings of the Convention of Conventional Weapons (CCW) Meeting of
Experts on Lethal Autonomous Weapons Systems (LAWS) held at Geneva
in April 2016. The primary issues debated were whether a pre-emptive ban
is justified and what should be the way forward for autonomous weapons in
the international arena. India has projected a need for increased systemic
control over international armed conflict in a manner that does not widen
the technology gap amongst states or encourage the increased resort to
military force in the expectation of fewer casualties or that use of force can
be shielded from the dictates of public conscience. India has also
highlighted the issue of international security in the case of the proliferation
of such weapon systems, arguably to non-state actors. India has noted that
there is a wide divergence on the key issues of definition and "mapping
autonomy" and that there is a need to resolve these issues.13 India has
emphasised the fact that such technology has both peaceful and
military applications, and that the CCW will be a relevant and
acceptable framework for addressing any issues of concern in future.

 Flexibility The Indian armed forces are still evolving in terms of


integrated procedures and data infusion. However, doctrines for the
integrated nature of warfare in the future are made, but not time-tested.
11
The Global Race for Artificial Intelligence: Weighing Benefits and Risks, IDSA Issue Brief, 23 Feb
2018 by Munish Sharma.
12
India and the challenge of autonomous weapons r. Shashankreddy.
13
Ibid.
117
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

This will result in inconsistent logic in algorithms, which eventually form the
backbone of any AI-based system. Whenever ready, this backbone will
decide the exact timelines for application. Also, the nature of rapid
changes in tactics in armed forces owing to hybrid threats is one
thing which needs to be kept in mind to avoid incorrect decision
making by these AI systems if implemented.

 Dependence on Foreign Technology The military may need to


adjust the acquisition process to match timelines and progress in other
countries and the civil domain. The availability of a valid and correct
database is a must for any application of AI. Procuring this technology from
any foreign country means a compromise of this database. If a solution is
chosen to use dummy data for application development, it will still lead to
teething problems for such a complex technology, and reliability will always
be in question. Also, our defence acquisition processes are not agile
enough for fast-paced development in AI, which will affect its war fighting
capability. Considering the progress of AI development in India vis-a-vis the
development of other nations, India is way behind the race. This will cause
Indian military equipment to become obsolete more quickly, despite not
conforming to conventional definitions of aging. Thus, perforce, we may
have to resort to foreign technology, as we have seen in the past
decade of military procurement in India. This dependence will only
deter the Indian military from asserting its will when desired and has
an inherent vulnerability to information.

 Challenges in Development of AI for Military From the Cold


War era until recently, most major defense-related technologies were first
developed by government-directed programmes and later spread to the
commercial sector. Examples include nuclear technology, the Global
Positioning System (GPS), and the internet. In contrast, civilian companies
are leading AI development, with the military domain adapting their tools for
national security functions. Today, AI development in the commercial
sector is leading the military sector. This reversal of the tradition of
strategic technology development is a concern. In India, the
development is still in a very nascent stage and military applications
are done piecemeal.

 Software Error/Lack of Subjectivity During the Cold War, and


each side was capable of a nuclear strike. An example to explain the
subjectivity involved in decision making is illustrated in the following
paragraph.

Lieutenant Colonel Stanislav Petrov was on a night duty in


bunker outside Moscow and his responsibility was to report the
missile launch up the chain of command to his superiors. In the
bunker, sirens blared, and a giant red backlit screen flashed
118
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

“launch,” warning him of the detected missile, but Petrov was


uncertain. Why would the United States launch only five missiles? It
didn’t make sense. A real surprise attack would be massive, an
overwhelming strike to wipe out Soviet strategic targets. Petrov was
not convinced the attack was real. But he wasn’t certain it was a
false alarm, either. He was right and there was no attack. Sunlight
reflecting off cloud tops had triggered a false alarm in Soviet
satellites. Humanity was saved from potential Armageddon by a
human “in the loop.” The AI machine would have done whatever it
was programmed to do, without ever understanding the
consequences of its actions.

Some decisions in conflict situations are straightforward.


Sometimes the enemy is easily identified, and the target is clear. Some
decisions, however, require understanding the broader context. The Indian
armed forces have yet to mature enough to make these artificial
intelligence-based decisions. Currently, every important military action
has the nod of higher authorities due to seamless communication in
each service. Changing the command structure and losing the
broader perspective due to AI will definitely take a long time to
employ in our environment.14

 Potential Weapons Arms Race The armed forces of the US and


China have already invested billions of dollars in developing LAW,
intending to gain strategic and tactical advantages over each other. This
runs the risk of an arms race. Similar to the support of chemists and
biologists for international agreements prohibiting chemical and biological
weapons, leading robotics and AI pioneers have called on the United
Nations to ban the development and use of LAWs. In the swelling
competition for AI technology progress, India fares average. Research
output from India in international journals ranks at 7, and the numbers
are one fifth of that of China.15 Given the predictable national security
value, some analysts subscribe to AI technology as a game changer.
Several analysts have cast the increased pace and magnitude of AI
development as a "Sputnik Moment" that may spark a global AI arms race.
Analysts warn that if military units rush to field the technology prior to
gaining a comprehensive understanding, they may incur a "technical debt."
The term "technical debt" means the effect of using AI systems will have
minimal risk individually but increase the danger of catastrophe as their
collective hazard is compounded by each new addition to the inventory.
This situation may be further exacerbated if nations engage in an AI arms
race.16
14
S Paul, Army of None, 2018, Ch 15, Pp
15
Sharma Munish,The Global Race for Artificial Intelligence: Weighing Benefits and Risks, IDSA
Issue Brief, 23 Feb 2018
16
HoadleyD,Lucas N, Artificial Intelligence and National Security,April 2018, Pp 17.
119
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

 Rapid Obsolescence of Infrastructure Ultimate success with AI


will likely depend on how suitable its environment is for such powerful
applications. While cloud computing is emerging as a major resource for
data-intensive AI workloads, military apparatus must continue to rely on
their existing IT environments for these projects due to budget
constraints.17 Considering the recent development of the IT
infrastructure of the Indian Armed Forces towards NCW, the
possibility of repeated up-gradation in the near future does not seem
feasible. The restrictions on upgrading processing and networking
infrastructure in the Indian Armed Forces are not only from restricted
budgets for procurement but also from the capability of each service to
adapt to new systems with artificial intelligence.

 Infrastructure Vulnerability AI algorithms are efficient only if they


are trained on large sets of data. The dataset may be too small,
mislabelled, inaccurate, or outright falsified by malicious actors. Big data is
a big problem if that data is bad. Data errors are a sensitive problem in the
national security sector, where the chief threat is not routine cyber-crimes
but sophisticated and well-funded nation-states. If a technologically
capable adversary is aware of the dataset your AI is training on, they will
have an edge in determining how to beat it. The availability of enough good
data is difficult. As a result, a lot of datasets are widely shared for
development. The worst-case scenario will be if the enemy inserts false
data into our AI systems. The AI learns the untrue version of reality that the
enemy desires. This situation will be very dangerous, especially because
the victim will not realise that their own AI data set is incorrect and may
continue trusting the incorrect results. The inner workings of AI algorithms
are particularly complex and unpredictable, even to their designers.
Examples of such a disaster exist in civil applications of AI. One such
example is that the Twitter feeds for Microsoft‟s Tay chatbot trained it to
send out racist content within hours of coming online. 18 In the recent past,
the Indian Army website and the Ministry of Defence website were both
hacked.19No single agency in India is charged with ensuring cyber and IT
security. According to data released by the Computer Emergency
Response Team-India (CERT-IN), 90, 119, 252 and 219 government
websites were defaced by various hacker groups in the years 2008, 2009,
2010 and January-October 2011, respectively. On July 12, 2012, in one of
the biggest cyber-attack on the India‟s official computer networks, over
100,000 e-mail addresses of top government officials were hacked in a

17
https://searchenterpriseai.techtarget.com/feature/Designing-and-building-artificial-intelligence-
infrastructure accessed on 10 December 2018.
18
https://www.theguardian.com/technology/2016/mar/24/tay-microsofts-ai-chatbot-gets-a-crash-
course-in-racism-from-twitter accessed on 25 December 2018.
19
https://www.firstpost.com/tech/news-analysis/indian-army-site-hacked-does-india-have-the-right-
attitude-to-tackle-cyber-crime-3666607.html accessed on 10 December 18.
120
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

single day.20 With lack of network integration, state of Cyber Security


apparatus of the nation and lack of valid AI trainable databases,
Indian Armed Forces and its systems have a considerable
vulnerability. Also, majority of the data sets currently available with
each service are not integrated or labelled nor are digitised in most of
the cases.

 Training and Acceptability Militaries which will utilise AI


technology must employ soldiers who have training in coding and who
understand how algorithms work. Similarly, at the user end, comprehension
of a complete picture or subjectively analysing results or solutions by an AI
system needs training and understanding of how an AI system works.
Currently, only a broad understanding of AI exists in the Indian Armed
Forces. If an AI-based system is presented without adequate training
and education about the changes that it is about to bring, then
applications of AI may have adverse effects. Also, from the point of view
of security and generating valid data sets, a huge amount of groundwork
needs to be done prior to implementation. This process may also demand
awareness from personnel and reasonable acceptance of non-human
decision-making.

Framework for AI in the Military

Human intelligence generally follows a sequence known as the perception–


cognition–action information processing loop, in which individuals perceive
something in the world around them, think about what to do, and then, once they
have weighed up the options, make a decision to act. AI is programmed to do
something similar, in that a computer senses the world around it and then
processes the incoming information through optimization and verification
algorithms, with a choice of action made in a fashion similar to that of humans.
Figure 2 illustrates how an autonomous system embedded with AI „thinks‟ and
makes decisions in this way. While there are many parallels between human
intelligence and AI, there are stark differences too.

20
Cyber Security: Avoiding a 2020 Pearl Harbour GurmeetKanwal CLAWS Journal.
121
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

Figure 1

Every autonomous system that interacts in a dynamic environment must


build and constantly update a world model (as shown in Figure 2). This means that
the world must be perceived (or sensed through cameras, microphones, and/or
tactile sensors) and then reconstructed in such a way that the computer „brain‟ has
an effective and updated model of the world it is in before it can make decisions.
The fidelity of the world model and the timeliness of its updates are the keys to an
effective autonomous system.

Autonomous UAV navigation, for example, is relatively straightforward


since the world model according to which it operates consists simply of maps that
indicate preferred routes, height obstacles, and no-fly zones. Radars augment this
model in real time by indicating which altitudes are clear of obstacles. GPS
coordinates convey to the UAV where it needs to go, with the overarching goal of
the GPS coordinate plan being not to take the aircraft into a no-fly zone or cause it
to collide with an obstacle. In comparison, navigation for driverless cars is much
more difficult. Cars not only need similar mapping abilities, but they must also
understand where all nearby vehicles, pedestrians, and cyclists are, and where all
of them are going in the next few seconds. Driverless cars (and some drones) do
this through a combination of sensors like LIDAR (Light Detection and Ranging),
traditional radars, and stereoscopic computer vision. Thus, the world model of a
driverless car is much more advanced than that of a typical UAV, reflecting the
complexity of the operating environment. A driverless car computer is required to
track all the dynamics of all nearby vehicles and obstacles, constantly compute all
possible points of intersection, and then estimate how it thinks traffic is going to
behave in order to make a decision to act. Indeed, this form of estimating or
guessing what other drivers will do is a key component of how humans drive, but
122
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

humans do this with little cognitive effort. It takes a computer significant


computation power to keep track of all these variables while also trying to maintain
and update its current world model. Given this immense problem of computation,
in order to maintain safe execution times for action, a driverless car will make best
guesses based on probabilistic distributions. In effect, therefore, the car is
guessing which path or action is best, given some sort of confidence interval. The
best operating conditions for autonomous systems are those that promote a high-
fidelity world model with low environmental uncertainty. These various factors
which will dictate the advent of AI in the military are discussed in the succeeding
paras.21

 AI Data Storage One of the biggest considerations is AI data


storage, specifically the ability to scale storage as the volume of data
grows. As organisations prepare enterprise AI strategies and build the
necessary infrastructure, storage must be a top priority. That includes
ensuring the proper storage capacity and reliability to deal with the massive
data amounts required for effective AI. The type of storage an organisation
may need depends on many factors, which include the level of AI an
organisation plans to use and whether they need to make real-time
decisions. Another factor is the nature of the source data. AI applications
depend on source data, so an organisation needs to know where the
source data resides and how AI applications will use it. Applications which
require analysing sensor data in real time or using post-processing will also
be critical to deciding the storage type. As databases grow over time,
organisations need to monitor capacity and plan for expansion as
needed.22

 AI Networking Infrastructure Networking is another key


component of an artificial intelligence infrastructure. To provide the high
efficiency at scale required to support AI, organisations will likely need to
upgrade their networks. Learning AI algorithms are highly dependent on
communications, and enterprise networks will need to keep stride with
demand as AI efforts expand. AI systems will require high-bandwidth and
creative architectures. As a preparation for AI, any organisation should
automate wherever possible to generate more data for the AI learning
process.23

 Processing AI Data Artificial intelligence infrastructure depends


on sufficient computing resources, which include Central processing Unit
(CPU) and Graphics processing Unit (GPU). A CPU-based environment

21
Cummings ML, Artificial Intelligence and Future of warfare, Chatham House, The Royal Institute of
International Affairs,pp3,4,5.
22
https://searchenterpriseai.techtarget.com/feature/Designing-and-building-artificial-intelligence-
infrastructure accessed on 01 January 2019.
23
https://searchenterpriseai.techtarget.com/feature/Designing-and-building-artificial-intelligence-
infrastructure accessed on 01 January 2019..
123
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

can handle basic AI workloads, but complex applications, such as for the
military, involve multiple large data sets and deploying scalable AI
algorithms. CPUs are best at handling single, more complex calculations
sequentially, while GPUs are better at handling multiple but simpler
calculations in parallel. To provide the necessary compute capabilities,
companies will have to resort to GPUs. Deploying GPUs will enable
organisations to optimise their data centre infrastructure and gain power
efficiency. Nvidia and Intel are both pushing AI-focused GPUs.24

 Preparing AI Data Organizations must plan the locations of stored


data, the movement of data across networks, and how processing will take
place. They also have to choose how they will prepare the data for use in
AI applications. One of the critical steps for successful enterprise AI is data
cleansing. Also called data scrubbing, it's the process of updating or
removing data from a database that is inaccurate, incomplete, improperly
formatted, or duplicated. Any company, but particularly those in data-driven
sectors, should consider deploying automated data cleansing tools to
assess data for errors using rules or algorithms. Data quality is especially
critical with AI. If the data feeding AI systems is inaccurate or out of date,
the output and any related business decisions will also be inaccurate.

 AI Data Management and Governance Another important factor


is data access. An organisation must have the proper mechanisms in place
to deliver data in a secure and efficient manner to the required users. A
data management strategy needs to ensure that human users and
machines have easy and fast access to data across a variety of endpoints,
including mobile devices via wireless networks. Access also raises a
number of privacy and security issues, so data access controls are
important. We need to look at technologies such as identity and access
management and data encryption tools as part of their data management
and governance strategies.

 AI and IoT IoT involves gathering and analyzing data from


countless devices, products, sensors, assets, locations, vehicles, etc., that
are connected via the internet.AI and IoT are closely tied because
organisations need to apply intelligence to gain insights from all the
information coming in from connected things. A staggering amount of data
can be generated by connected objects, and it will be up to organisations to
integrate, manage, and secure all of this information. From an artificial
intelligence infrastructure point of view, companies need to look at their
networks, data storage, data analytics, and security platforms to make sure
they can effectively handle the growth of their IoT ecosystems. That

24
Ibid.
124
PINNACLE 2022
Artificial Intelligence and Future of Warfare: Roadmap for the Indian Army

includes data generated by their own devices, as well as those of their


supply chain partners.25

 AI Training Training and skills development are vital for any IT


endeavour, and especially enterprise AI initiatives. Organisations will need
data analysts, data scientists, developers, cyber security experts, network
engineers, and IT professionals with a variety of skills to build and maintain
their infrastructure to support AI and use artificial intelligence
technologies.26

Conclusion

With the present rate of development of AI technologies would wide and


the race for achieving dominating standards in the same by military powers of the
world, this concept is going to revolutionise various domains of military. Increased
risk-taking ability and the priority of force preservation provide a further push
towards progress in this field. However, it has also been realised during various
trials and live situations on the ground that the current generation of AI equipment
is far from being operated without human supervision. Hence, the future is likely to
witness a quantum leap in the utilisation of AI-based equipment though it is still
controlled/directed by a human leader/boss.

25
https://searchenterpriseai.techtarget.com/feature/Designing-and-building-artificial-intelligence-
infrastructure accessed on 01 January 2019.
26
Ibid.
125
PINNACLE 2022
ARTIFICIAL INTELLIGENCE AND THE
INTELLIGENTIZED WARFARE
- Lieutenant Colonel Manu Joseph Chacko

Introduction

The military architecture of future will consist of a new array of sea, ground,
and space-based sensors; unmanned combat aerial vehicles (UCAV); and missile
defence technologies. Military forces will be agile, and attacker will take advantage
by operating faster than a defender can observe, orient, decide how to respond,
and act on that decision. The attacker will thus place himself inside the defenders'
Observe, Orient, Decide, and Act (OODA) loop, destroying an adversary’s ability
to conduct an active defence. To execute the OODA process faster than the
enemy is the core concept of future digital and information warfare. Automated
systems, assisted by AI in some form or another, may be a way out of this
problem. The advances gained in the field of AI can be utilised by unmanned
systems to be able to assess operational and tactical situations and decide on an
appropriate action. Information will drive the success of command and control.
These systems will collect and analyse data and provide options to the
commander. The essential ingredients for development of AI are as listed below: -

 Data Data is an important foundation of intelligent warfare. Data is


considered the "new oil" 1and big data the "most important resource" in
intelligentised warfare. This is primarily because all the important functions
of AI will depend on the availability of data.

 Algorithms Futuristic warfare will revolve around algorithms. AI


with better algorithm will drive decision making in war. Algorithm is the core
element for transforming war from informatised to intelligentised.

 Computing Power Real-time computing of large amounts of data


is an essential component of intelligent warfare. Intelligent warfare cannot
be conducted without strong computing power. The next generation of
computers such as quantum computers, photonic computers, and
biocomputers will provide unprecedented computing potential and help
advance application of AI technology in warfare.

In simple terms, AI can be defined as teaching machines to learn, act, and


think as humans do. Machines are being endowed with the cognitive ability to think
and behave like humans. Broadly, it demonstrates association with HUMINT such
as planning, learning, reasoning, problem-solving, knowledge representation, and,
to an extent, social intelligence, and creativity. Machine learning, which is a subset

1
Data Is the New Oil - And That's A Good Thing by Kiran Bhageshpur, https://www.forbes.com/
sites/forbestechcouncil/2019/11/15

PINNACLE 2022 126


Artificial Intelligence and the Intelligentized Warfare

of AI, uses algorithms to analyse data and make intelligent decisions based on
what it has learned without being explicitly programmed. Machine learning
algorithms are formulated with large sets of data to suitably process, match, and
recognise patterns to come to a logical conclusion.

AI systems can be classified into following heads:-

 Artificial Narrow Intelligence It is also referred to as "Weak AI"


or "Narrow AI"2, which only has a narrow range of abilities. However, this is
the only type of AI that has been successfully applied till date. Various
applications, including facial recognition, speech recognition, driving a car,
and searching the internet (bot army), can be classified under this
category.

 Artificial General Intelligence Machines that can work at par with


human capacity are classified into this category. It can think, understand,
and act in a way that is indistinguishable from that of a human in any
situation. This is also termed "Strong AI" or "Deep AI".

 Artificial Super Intelligence3 This type of AI is capable of


outperforming humans It is where machines become self-aware and
surpass the capacity of human intelligence. In this intelligence system,
machine intelligence exceeds human intelligence qualitatively and
quantitatively.

Development of AI technology has elevated computers from computing,


storing, transmitting, and executing commands to thinking, reasoning and
replacing them by extending the functions of the human brain. Cross-domain
asymmetrical and unconventional fighting in battle will become the new normal.
There will be an integration of human and machine intelligence in "Intelligentised
Warfare". Combining wearable devices and gadgets implanted into human bodies,
humans and machines into brain-machine interfaces, and external skeletal
systems, will "comprehensively enhance the inherent cognitive and physiological
capacity of human fighters and will forge out superhuman combatants."

The present dimensions of the battle fields like land, sea, air, EM, and
cyber will be further upgraded to include a cognitive dimension in warfighting. AI
will reshape warfare in every dimension and within every realm. AI will avoid
human fighters being the first line of fighting, and the future of warfare will be
upgraded to "machine-on-human" or "machine-on-machine" warfighting. It is
believed and has been discussed that the human brain will not be able to keep
pace with the future complex and dynamic conflict scenarios, further necessitating

2
Narrow AI vs Artificial General Intelligence – the key difference and future of AI Narrow AI vs
Artificial General Intelligence - the key difference and future of AI - ai.nl 18 February 2022.
3
Ibid.
127
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

the emphasis on the utilisation of AI in decision making. The pursuit of military


advantage through AI creates not only the capability but also new vulnerabilities
that could exaggerate the risk of momentum-driven escalation, particularly in the
cyber domain. There is an inherent risk and reliance on complex automated
systems in which errors and malfunctions are not only probable but probably
inevitable.

THE INTELLIGENT BATTLEFIELD

With the prominence of AI and autonomous weapon systems in future


battles, there are likely to be profound changes in future battle scenarios. AI and
autonomous weapon systems will be the dominant forces in the battlefield. The
versatility that unmanned weapon systems possess and the removal of humans
from the battlefield will further expand the battlespace. Removing humans will not
only entail a reduction in the size of platforms but also reduce the number of
weapon platforms. The future unmanned systems will automatically detect targets
and launch independently based on the location, size, and state of the target.

With AI being a potent enabler, it has endless possibilities in the military


sphere. In the military sphere, AI can be utilised for intelligence, information and
data analysis and distribution, realistic war gaming, prediction, training simulations,
communication, logistics, movement etc. The effective utilisation of AI in intelligent
battle fields is described in the succeeding paras.

 Shortening Own OODA Loop AI can be effectively utilised in the


surveillance grid to provide seamless integration of various sensors and
platforms. It would assist in the formulation of a grand picture at the
strategic level to further fine-tune the situation and present it to field
commanders. It would assist in analysing the situation and generating the
courses of action necessary to exercise effective command and control. It
would assist in decision making by creating reliability indices for various AI-
generated courses of action and could also predict future events for
various courses to certain degrees. It could use prediction and pattern
recognition to fill in the gaps resulting from the fog of war and further
shorten its own OODA loop, thereby ensuring effective decision making.

 Force Multiplier The intelligentised battlefield would provide a


scenario of seamless integration of combat elements like soldiers, vehicles,
weapon systems, aircrafts, ships, submarines, drones, unmanned vehicles,
vessels, etc. through one single information base system. It would then
optimise the data being sent to each individual entity on the battlefield,
attuned to his or her role and requirements. Such integration and effort
coordination would have an immense force multiplier effect, which would
be an inescapable necessity for the success of operations. Future wars are
likely to be more information-centric and data-intensive. AI would not only
ensure well-coordinated military action, resource allocation, movement,
128
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

and administration, it would also enhance flexibility in switching of forces to


cater for any unforeseen contingency.

 Enabler for Multi Domain War Future conflicts are likely to be


undeclared, multi-domain, and subversive, involving the incorporation of
various hybrid, strategic, and grey zone situations where military actions
will take place in complex and densely populated environments. The
adversaries' use of unconventional methods, such as targeting groups of
people of a specific ethnic base, leaders, and vital installations with
unconventional means and weapons such as drones and cyber warfare, is
predicted to be how future wars will unfold. The artificial intelligence
systems would be the functional enablers for the forces to operate in the
future multi-domain hybrid environment.

 Increased Implementation Narrow intelligence has been adopted


by world armies in modern combat battlefield elements such as fighter
aircraft, Unmanned Aerial Vehicles (UAV), naval battle systems, marine
crafts, battle tanks, missile systems, and transport systems. It is being used
to some extent for on-board system integration, optimization, sensor
fusion, and even human-triggered weapon launch. AI in the present
scenario is in a nascent stage of being adopted across the globe. In time to
come, as AI improves, it will vastly improve and add to the combat
capability and survivability of military systems.

 Autonomous Weapons Systems Intelligent warfare will not only


improve battle management but will also be used effectively in all major
offensive and defensive weapon systems of the world armies. In the case
of weapon systems, AI would be utilised in decision-making and in aiding
decisions for weapon launch. Autonomous weapons, by various scholars,
are already being described as the third revolution of warfare, after
gunpowder and nuclear weapons. The AI robotics system's being
dovetailed with soldiers in the first wave of attack, would scale down the
threat to the assaulting forces. The AI system would also enable effects-
based operations (EBO) by intelligently targeting selected people and
installations and creating the desired effect with minimal collateral damage.

 Reduce Expenditure to Exchequer The future AI military


systems would reduce the need for maintenance of regular forces, thereby
ensuring downsizing of militaries. Over the years, military expenditure has
been increasing exponentially for all the developed nations. This is
primarily because most of the systems, including weapon platforms, are
still human controlled and only a few are completely automated. AI
unmanned systems, requiring fewer inbuilt systems, would bring much
more economy to military operations as compared to manned systems. A
manned F-22 fighter aircraft, for example, is approximately 18 times more

129
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

expensive than an unmanned Predator Drone4. PLA is already in the


process of downsizing its military and thereby expanding its tech base.

Military environments are chaotic, unpredictable, and adversarial. Even


though AI can perform admirably in some situations, it can suffer significantly in
others. AI systems used by militaries may suffer accidents and be manipulated by
adversaries. The limitations associated with AI are as under: -

 Uncertainty in Battlefield The present and future battlefields are


riddled with various distortions like smoke, dust, etc. There is also a
possibility of an adversarial attack corrupting the data to deceive the AI,
which may further impact the surveillance, weapon grid system, and other
operations. Uncertainties for AI from the environment could be many others
too, including sensor, component failures or adversarial attacks, which may
lead to ambiguous results. Also AI may force decision makers into multiple
scenarios resulting in dissipation of resources.

 Limited Funding It is appreciated that military usage of AI would


be negative and hence, world armies are at times sceptical about the
development of the same for manufacturing various AI-based weapon
platforms. This is one of the reasons for the limited funding of various
governments towards defence forces' adoption of AI vis-à-vis their civil
counterparts. Limited funding is one of the major hindrances to carrying out
necessary R&D in AI.

 Acceptance of Lethal Autonomous Weapon System (LAWS)


There is a faction of experts and scientists who find it is a gross violation of
human ethics and are strongly opposed to the use of autonomous weapon
systems, which is further delaying the development of AI in many countries.
Task accomplishment using fewer humans and risk aversion would be the
significant advantages of LAWS.

 Hardware and Design Challenge For effective usage of any


product in the military, it must be accepted by "Military Standards", which
involves it having to withstand high stress, tough usage, and battlefield
ruggedness. The design thus involved should withstand rigorous testing to
deliver a much stricter performance guarantee, which usually takes more
time and thus delays the production.

 Inflexibility Based on current technology levels, AI can only solve


problems within a specific range and lacks the ability to respond effectively
to new situations. The strength of human intelligence is, thus, very

4
The US and its UAVs: A Cost-Benefit Analysis By Ashley Boyle on Jul 24, 2012 The US and its
UAVs: A Cost-Benefit Analysis American Security Project.

130
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

important in warfare where unpredictability and chaos are central elements.


In its present state of development, no AI system can replace the flexibility,
robustness, and generality of human intelligence.

 Trust Building Considering less human involvement during the


fog of war, when it is appreciated that data may get distorted and the
system may give erratic results, building trust in the AI weapon system is
one of the major challenges that is being encountered. The AI systems are
being run under various conditions to develop a fool proof mechanism to
avoid any failures.

ROADMAP FOR INDIAN ARMY

India is at a nascent stage towards adopting AI in both the defence and


civil sectors. However, by the latest report of International Data Corporation (IDC),
India Artificial Intelligence Market, 2021 that predicts India’s AI market to reach
USD 7.8 billion by 2025 at a compound annual growth rate (CAGR) of 20.2%, from
a market value of USD 3.1 billion in 2020. Moreover, the AI services market is set
to lead the overall AI market growth in India by 2025 at a CAGR of
35.8%5. According to a survey, a large no of Indian companies in private sector
are adopting AI. However, recruiting skilled professional already trained on AI
remains a challenge. Many analysts have already commented that India’s entry
into this field is "late". It has also been highlighted that the late entry will have a
serious ramification not only in defence but in the civilian sector too.

With our adversaries adopting AI-enabled tech in their defence sectors, it is


imperative that India should also adopt a methodology to imbibe AI for the
development of military technology. Following are certain implications that India
may face in future conflicts: -

 Targeted Attacks Our adversaries have already developed the


capability of AI-enabled drones capable of carrying out targeted attacks. In
the event of conflict, they may utilise these weapons to carry out targeted
attacks on high-profile individuals to disrupt their own command and
control systems.

 War Experience The Indian Army lacks modern operational


training, and its lack of understanding of fog of war may lead to errors or
unrealistic expectations about the prospects for technology on the
battlefield. It may happen that Indian intent and capability may be
exaggerated, which may accidentally trigger unwarranted conflicts.

 Export of Technology It cannot be negated that China could


export its arsenal of autonomous weapons and AI technology to India’s

5
“India’s AI market to reach USD 7.8 billion by 2025,” says IDC’s latest report on AI (indiaai.gov.in)
131
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

other potential adversaries to subsequently engage India from more than


one front during the event of conflict. Furthermore, the availability of this
technology to non-state actors may further constrain the ability of
significant forces to carry out any conventional operations.

 Shortened War With autonomous weapons being used in future


wars, it is likely to be a shorter war. Political leaders are thus required to be
more mentally agile, robust, and able to make relevant and swift decisions
to avoid high casualties on either side.

 Inadequate Framework In the present scenario, there is no


framework that has been formulated to absorb intelligent warfare into our
own military doctrine vis-à-vis China, whose political leaders are already
boasting about their autonomous weapons arsenal and who, by 2030, will
be a leader in AI technology in the military domain.

Recommendations and Way Ahead

 Roadmap towards Adoption of AI High priority should be given


towards R&D of AI, AI-based environments, and AI-based devices. A
definite roadmap needs to be prepared towards adoption of AI and,
accordingly, budgeted.

 Accelerating Hardware and Software Production The "Make in


India" policy outlines the accelerated growth of hardware production in
India. It is recommended that adequate impetus be provided for chip
production to make the nation self-reliant. In addition, the requisite stimulus
must be provided for designing algorithms and software.

 AI in Educational Institutes It is imperative that an adequate AI


knowledge base be created in the country for future development. Hence,
there is an immediate need to introduce AI in educational institutes or
modify their educational curriculum to include AI while at the same time
taking policy initiatives and incentives for retaining AI talent within the
country.

 Keeping Abreast with Latest Technology Periodic technology


orientation programmes for the policy makers and the Indian Army are to
be organised to keep them abreast with the latest technological
advancements happening all across the world. It will give them ideas on
how to implement the same in their own organization.

 AI Based Wargames There is a requirement to develop an AI-


based integrated war management and battle control system at different
levels of priority. Aspects like scenario building, logistics, training,
132
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

intelligence, movement, communication, etc besides comb forces AI


system like virtual reality system, which would cut the expenditure on
peacetime training activities is highly recommended to be adopted in Indian
defence forces.

 Creating a Government Driven Public Organisation for AI To


raise public awareness about AI and its importance across sectors, India
must establish an organisation like the Chinese National Innovation
Institute of Defence Technology under the aegis of a National University for
Defence Technology.

 Expanding AI Defence R&D While DRDO runs the Centre for


Artificial Intelligence and Research (CAIR)6 focused on the development of
AI for defence, its adoption of AI and AI-enabled systems is still in its
nascent stage. So far, the DRDO has developed a few unmanned systems,
one visible example of which is the Mission UNmanned TRAcked called the
MUNTRA, which has three variants for recce, mine detection, and
surveillance.

 Awareness of Adversary’s AI Tech To mitigate the risks of


surprise, IA must continue to track and monitor new directions of our
adversary’s modernisation plans. India should improve its capacity to
leverage OSINT technology and reprioritize the targeting of collection
activities as necessary.

 Developing Effective Response Mechanism Adequate efforts


should be made to ensure safety of our own AI system against an
adversary’s potential threat. As the character of conflict evolves and
complexity increases, continued exploration of mechanisms for confidence
building and crisis management will become more critical.

 Improve Bilateral and Multilateral Relations India should


continue to pursue productive conversations with other global powers to
establish consensus on the legal and ethical parameters of employing
autonomous weapons during conflict. Best practises and technology may
be shared with other nations to form a comprehensive architecture.

Conclusion

AI-enabled technology has already started to become the driver of change


for mankind, and its effects are likely to intensify further in the future. This article
concludes with the observation that advances in artificial intelligence have ensured
that autonomy has now crossed a "tipping point". There is no alternative to

6
Centre for Artificial Intelligence & Robotics (CAIR) | Defence Research and Development
Organisation - DRDO, Ministry of Defence, Government of India
133
PINNACLE 2022
Artificial Intelligence and the Intelligentized Warfare

embracing the upcoming technology promptly and reviewing and adjusting the
policies to cater for the quick, successive changes that are likely to ensue in the
future, including in the defence sector.

With the realisation of the paramount importance of AI technology, a global


contest has already begun to achieve global leadership in the field. India’s entry
into the field of AI development and exploitation has already been delayed, and
considering its situation, there is a requirement to accord AI development an
immediate high priority to avoid suffering permanent disadvantage vis-a-vis others,
particularly in the Indian Army. The path forward is difficult, as there are numerous
challenges, such as adopting a forward-thinking approach, drafting policies and
roadmaps, developing, and retaining AI software skills, establishing an industrial
base for hardware, and enticing entrepreneurs to invest in AI software and
hardware development. The stirring has already begun, but there needs to be a
prodigious thrust to catch the required trajectory.

134
PINNACLE 2022
ARTIFICIAL INTELLIGENCE AND THE FUTURE OF
INDIAN MILITARY LOGISTICS
- Brigadier Achal Dilip Kumar

By far, the greatest danger of Artificial Intelligence is that people conclude


too early that they understand it.
- Eliezer Yudkowsky

Introduction

Circa 2035: A military convoy carrying troops and supplies is well on its
way to its destination. The convoy itself is a mix of human-driven vehicles and
Autonomous Vehicles (AV). The AV at the head of the convoy changes course/
route based on information received on the current route being untenable, other
vehicles follow. The convoy commander gets all the requisite information
simultaneously on the screen in his vehicle. The large convoy is split into two,
based on the developments in the fast-moving battle taking place necessitating
troop re-deployment and redistribution of supplies. Signal communications for the
same have been obviated as all changes (route deferment, re-deployment and re-
distribution of resources) have been executed through Internet of Things (IoT).
Stores at the depots in the hinterland are accordingly replenished based on these
changes, from the source prime vendor, seamlessly. Simultaneously, Intelligence,
Surveillance & Reconnaissance (ISR) UAVs and Cargo UAVs are operational in
order to both watch the emerging battlefield dynamics as well as to supply niche
small teams of Special Forces at their point of concentration. A sudden specific
logistics requirement that has emerged, has resulted in a smart contract being
triggered and executed with clearly laid down delivery times including locations.
Robots work incessantly sorting, stacking and identifying varied items in dual-use
warehouses, while loading robots and cobots (collaborative robots) ensure fast
loading of essential cargo vehicles leaving soon for their destinations. The entire
warehouse is overseen by just one Junior Commissioned Officer with one
assistant.

The above scenario may appear as a scene out of a sci-fi movie but
Artificial Intelligence (AI) is making such possibilities come within the realm of what
is implementable in the very near future. Some of these technologies are already
in use by leading firms like Amazon, Walmart etc. AI will soon touch every facet of
life, with warfare being no exception. Logistics and supply chain leads itself
easily to the adaption of AI, as many of the processes are already complex
(complex relationships between end-users/suppliers/manufacturers/transporters
etc) and use humongous volumes of data and data processing. Technologies like
AI-enabled blockchain will ensure security of logistics processes making them less
susceptible to manipulations or distortions. The need for asset visibility too has
exponentially increased to facilitate better Decision Making (DM). In the Armed
Forces too, these complexities will manifest further as we increase our adaption of
PINNACLE 2022 135
Artificial Intelligence and the Future of Indian Military Logistics

the logistics and supply chain of the civil sector. Hence, a seamless interface
would speed the processes at both ends. The answer lies not only on the use of
technology but specifically AI, in order to tackle these challenges without breaking
a sweat.

World over major armies have attributed great importance to military


logistics for applications of AI. A case in point, out of seven application areas for
which the Chinese military is awarding AI-related equipment contracts ie
intelligence analysis, information warfare, autonomous vehicles, logistics,
training, command and control, and target recognition, two that are in the realm of
logistics.

For the purposes of this article, the scope of logistics includes Army Supply
Corps (ASC), Army Ordnance Corps (AOC), Corps of Electronics and Mechanical
Engineers (EME) as also Engineer and Medical stores, since they lend themselves
rather easily to benefit from advances in the field of logistics. The minor logistics
services have been excluded.

What is AI?

The Britannica defines AI as the ability of a digital computer or computer-


controlled robot to perform tasks commonly associated with intelligent beings.
Oxford Languages defines AI as the theory and development of computer systems
able to perform tasks that normally would require human intelligence. Eg visual
perception, speech recognition, decision making and translation of languages.

A word of caution is to avoid confusing prevalent cutting-edge technology


with AI, while AI and Machine Learning (ML) can boost the efficacy of all these
technologies if used appropriately.

136
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

Themes for AI in Indian Military Logistics

Supplies/ Technology
Planning Transportation Functioning Back office Training
Stores

Smart Automated
Demand Autonomous Transparency &
Warehousing with Document Big data LMS
Forecasting vehicles Decision Making
Robots Processing

Automate Manual
Supply Planning Damage detection FMS, GPS Customer care Cloud computing Robots for training
Office Tasks

Predictive Uberization of Customer Service


Decision making IoT AR, VR, IVR
Maintenance transport Chatbots

Predictive Prime vendor 5G


Drones Metaverse
capabilities programme communications

Intelligentization of Self Driving Swarm


Dynamic pricing
logistics support Vehicles intelligence

Suply Chain 4.0 Smart contracting Last mile solutions

Metaverse convoy
Risk Management
planning

Route
Optimisation

Figure 1: Themes for AI

The DHL Logistics Trends Radar 2020 lists AI as a „High Impact‟ trend in
the time horizon of less than five years as shown below.

Figure 2: Logistics Trend Radar

137
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

In the coming years, the fusion between military logistics efforts and civil
logistics infrastructure and supply chain, will form an important component for
overall mission accomplishment facilitated by a seamless support system.

Certain characteristics of AI which empower and enhance specific


technologies (features), which bear potential to revolutionise supplies, are listed
below.

Characteristics Features
• Transparency & Decision -Making • Quality control
• Contextual Intelligence • Supply Chain 4.0
• Predictive capabilities • Blockchain, big data, IoT, cloud computing, 5G
• Generic optimization algorithms for faster delivery, lower commuications
costs; solve logistics problems • Efficient warehousing, robots, cobots
• Empower smart logistics systems • Intelligent DM support
• Risk management in supply chain, logistics flow • Uberization of transport, Fleet Management Systems (FMS),
optimisation at situations of crisis AV, drones
• Intelligentization of logistics • 3D/4D printing, additive manufacturing
• Logistics network orchestration • Swarm intelliegence
• Teeth-to-Tail Ratio (T3R) • Chatbots for customer support
• AI to tackle cyber threats to logistics
• Back office automation

Figure 3: Key Areas for Impact through AI


Intelligentisation of Logistics Support

The modernisation of many armies is centred around an integrated


approach to technological development which is also proportionately applied to
logistics support. Beyond the battlefield, AI is expected to contribute to
transformation resulting in the “intelligentization” of logistics support, over which,
AI will leave its indelible mark. The constituents of this intelligentization are
summarised below:-

 Supply Chain 4.0

 As the Internet moves from Web 1.0 to 2.0 and now 3.0, it
empowers technologies like AI to touch any facet of logistics
support. AI would significantly empower smart logistics/ supply
systems, which along with an integrated supply chain, constitute
Supply Chain (SC) 4.0. The four layers of a smart logistics system
(all of which would get impacted through AI) are: (a) perception
layer, (b) transport layer, (c) storage layer, and (d) service layer.
The technologies that underpin SC 4.0 include blockchain, smart
contracts, applications of AI, cyber physical systems (CPS), IoT etc.

138
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

 Logistics network orchestration will become common-place,


ie, harmonization of physical and digital networks on which logistics
services ride, using AI technologies. AI, big data and deep learning
may minimise the manual input needed to move from less reliable,
automated systems to more accurate, autonomous systems.
Creating a „nervous system‟ for military logistics through SC 4.0, will
take up each of the existing limitations of the logistics support
system and transform it into enhanced „avatars,‟ as represented in
Figure 4.

• Limited Asset Visibility Supply Chain • Full asset visibility


• Manual Fleet Management,
driver fatigue 4.0 • Smart transportation, fleet
management systems,
• Piecemeal ERP autonomous vehicles
• Lack of Jointness • ERP common to logistics of all
• Limited Warehousing three Services
• Traditional Contracts • Jointness
• Manual Invoicing/ Inventories • Integrated • Robots, Cobots, Automated
• Forecasting through SRDS supply chain + Warehousing
• Time consuming Supply Chain • Auotmated work flow
smart logistics • Predictive forecasting & demands
• Repairs and recoveries
through traditional methods =Supply • Faster delivery, reduced costs
• Limited exploitation of Big • Chain 4.0 • 3D, 4D printing, Additive
Data, AI, ML, IoT, Cloud Manufacturing
Computing, 4G/5G networks & • Exploitation of all technological
wireless communications for tools
logistics • Threat of cyber attacks is more
• Limited cyber security threats but can be tackled through AI
• Traditional training methods • Smart training - training robots,
learning management systems

Existing Empowered

Figure 4: Advantages of Supply Chain 4.0

 Big Data Logistics involves handling phenomenally large volume


data. As a strategic resource, big data is the basis of building a smart
logistics system, created through a process of storage, transportation,
distribution and processing. AI will effectively ensure the processing of big
data as also assure its security. Cloud computing will also be the norm to
render processing abilities, which are scalable.

 IoT It is argued that logistics should be the first application field for
IoT, where the possibilities of its employment are endless. It would enable
hardware, vehicles and equipment to communicate with each other. In

139
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

transport, it facilitates the ability to diagnose potential defects and take


preventive measures in time. It would also ensure orders for spares are
placed with Just-in-Time (JIT) so that there is no lag time and no
unnecessary stocking.

 Use of 5G Communications and Wireless Networks 2022 will


witness the introduction of 5G communications in India which will
revolutionise many other technologies like AI, AV, IoT etc. This will also
bring in its own set of cyber security issues which will need to be
addressed.

 Swarm Intelligence in Logistics Military logistics thus far, was


dependent upon and implemented as a top-down hierarchical model.
Swarm intelligence will transform logistics by reversing engineering rules
from observed behaviour. Swarm logistics will facilitate decentralised
Decision Making (DM) without compromising accuracy and protocols. Such
outcomes could be useful in a CI/CT environment and in military operations
other than war (MOOTW).

 Contextual Intelligence AI/ML offer logistics and supply chain


professionals with contextual information which can be used to manage
inventory better, reduce operating costs, responding to end-users quickly,
shorten delivery periods, mitigate risks, boost productivity, and enhance
customer delight. AI thereby impacts various areas of warehousing
management, logistics and supply chain management significantly.

 Intelligent DM Support Knowledge-based Decision Support


Systems (DSS) use AI for DM, due to its ability to gather and analyse data,
identify problems from these data, and finally find and evaluate solutions.
Such DM is more efficient and faster.

As opposed to the current system of basing procurement and


logistics decisions on manual reports and returns and a system of SRDS,
AI will enable instant DM and execution/ implementation. Procurements
would be based on actual need and hence wastages, over-stocking or
stocking at the wrong locations would be avoided. The need for, inter-
command transfers would be reduced, cutting costs. DM vis-à-vis
employment of logistics resources like convoys, transport, repair and
recovery etc would be streamlined, just to name a few applications.

AI: Impact on the Paradigms of Supplies

 Predictive Logistics, Automated Planning and Scheduling In


supplies, the ability to correctly forecast and place demands as also
planning and scheduling various supply-related tasks would get automated

140
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

reducing errors. The delicate balance between JIT and Just-in-Case (JIC)
stocking and distribution would be resolved through AI tools.

 Smart Contracting Blockchain and AI will facilitate smart


contracts, which are programmed to execute at the instant when conditions
are most appropriate to the organization as well as when most required.
Generally, smart contracts are signed by relevant parties, and then
attached to blockchain data in the form of program codes. They are then
recorded into specific blocks in the blockchain after network transmission
and node verification. The blockchain can monitor the status of smart
contracts in real time, and activate and execute the contract after checking
external data sources and confirming that certain trigger conditions are
met. Smart contracts can ensure traceability, irreversibility and transaction
security in the absence of third parties.

This will greatly reduce paperwork and errors of omission and


smoothen the supply process. It will of course, ride on the significantly
enhanced infrastructure in the country that permits the sourcing from areas
earlier considered geographically untenable, thereby bringing more
markets within reach and standardising quality.

 Prime Vendor Programme As of now, most supply depots in the


Armed Forces are dealing with multiple agencies, with separate contracts
for meat products, dairy products, fresh, water, beverages etc, while there
are central procured items contracted centrally by Army Headquarters for
dry items, edible oil etc. This system carries with it challenges of
coordination, addressing customer satisfaction etc. A concept of „Prime
Vendor‟ if implemented would ensure that a sufficiently reputed and large
vendor can undertake to supply 100% of the requirement of supplies
(perishables and dry supplies) as well as Fuel Oil and Lubricants (FOL)
products for a station at the doorstep of a receiving unit. This would reduce
the need to maintain large establishments of supply depots and also
prevent unnecessary move of military transport for collection of supplies. AI
will play a part in accurate estimation of the dynamic requirement and also
facilitate dynamic pricing discussed subsequently.

 Optimisation Algorithms AI/ML would render optimization


algorithms that would allow for instant resolution of logistics and supply
chain problems that crop up from time to time. For instance, the information
that a certain warehouse or depot has been rendered non-operational due
to enemy action or other reasons, would trigger the algorithms to
automatically re-route to other depots on the quantities necessitated. A few
vehicles being rendered out of action may trigger action to re-allocate
transport from other sources or formations.

141
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

 Dynamic Pricing AI has enabled the geographic spread of the


distribution system. A common bugbear in getting the bigger players to
participate in the supply process has been the unviability of contract rates
as it is binding for a complete year whereas the retail chains work on a
daily or weekly pricing model (which, from vendor‟s perspective, is
financially risky, inhibiting participation in contract processes). As AI
proliferates, and with resultant transparency in pricing, adaption of a
dynamic pricing model is tenable, which would undoubtedly attract the best
retail chains to participate in the contracting process.

Transportation Paradigms

 Autonomous vehicles Self-Driving Vehicles (SDV) may obviate


several manpower issues in transport management. It will change the face
of convoys as cargo vehicles will be able to operate 24/7 without need for
rest of drivers. SDV would allow for radical restructuring of units due to the
redundancy of MT drivers. Absences for leave, sickness, courses etc would
become a thing of the past. It would further facilitate the implementation of
measures for a more favourable teeth-to-tail ratio. SDV are also expected
to be safer than human-driven vehicles as the technology matures.

 UAVs and Last-Mile Connectivity Solutions As Animal


Transport (AT) units get disbanded in a phased manner, the expediency
exists to mitigate the absence of AT, as infrastructure improvement alone is
not likely to solve the problem in the near future. Hence, the need for last-
mile connectivity solutions like cargo UAVs. These UAVs will drastically
reduce air maintenance costs in remote and high-altitude locations.
Concepts like MULE (Multi Utility Logistics Equipment) will also be useful
as load carriers in difficult terrain where AT used to operate.

 Fleet Management Solutions (FMS) Transportation will undergo


a revolution. While FMS already exists, AI will further enhance capabilities
of such systems giving staff at all levels full control over this resource and
facilitating DM in their optimal employment. „Uberization’ will be the norm,
a term which encompasses cooperative intelligent transport systems,
intelligent route optimization offering dynamic routing facilities, last-mile
delivery using UAVs or Unmanned Ground Vehicles (UGVs), use of
autonomous trucks/ fleets, truck platooning or caravanning of groups of
semi-trucks, etc and all of these will bank on an AI backbone.

 Metaverse Convoy Planning Facebook being renamed as „Meta‟


recently highlights the importance the world of technology places on the
potential of the virtual world (the „Metaverse‟). Undoubtedly this technology
would be harnessed and enhanced with AI for applications in supply chain
and logistics support systems. It is not difficult to visualize numerous
military logistics applications as the technology matures. For instance,
142
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

convoy planning can be done by the senior „Q‟ staff in the metaverse which
can fully and safely replicate the routes, terrain and operational conditions.

Warehousing Technologies

Figure 5: Container Unloader Robot, Warehouse Robot, Cobot

Warehousing technology has grown in leaps and bounds in recent years at


least in the civil sector. There is a need for this to percolate to military too.
Companies like Amazon have vast warehouses (fulfilment centres) run by very
minimal manpower due to AI-based automation, robots, cobots and automated
loading/ unloading of vehicles. Cobots can assist combatants or logistics
personnel in carrying out repetitive tasks.

These aspects, if properly exploited, could favourably impact Teeth-to-Tail


Ratio. AI has been shown to improve logistics losses by 15%, inventory levels by
35% and service levels by 65%, thereby significantly improving the efficiencies of
the logistics processes.

Repairs and Recovery Support

Additive manufacturing (3D, 4D printing) technologies will greatly enhance


repair and recovery support by obviating the spares problem, by making available
the desired spares close to the point of application. Lengthy lines of
communication could thus be neutralised/ circumvented transcending
geographical limitations. In June 2017, the United States Navy exploited the
blockchain technology to improve the security of additive manufacturing systems.
They recorded the entire process of component design, prototype manufacturing,
testing, production and final processing, so that users could look into any specific
data, and provide alerts in case of component damage or at the end of its lifecycle.

Blockchain and AI for Transparency and Decision Making

Transparency is a pre-requisite of the logistic echelons and this would be


greatly enhanced through AI. Consequently, even the DM process would be
expedited in an atmosphere of technology-enabled trust, as clarified in Figure 6
below.

143
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

Data-driven AI/ ML requires high integrity data for their use in AI functions
and requires significant amounts of training data from diverse sources including
Internet of Things (IoT) devices/ sensors, which will be facilitated by blockchains,
which thus offer significant data management benefits.

Blockchain (also called distributed ledger technology), is a user community-


managed ledger technology that ensures transmission and access security with
cryptography, enables consistent data storage, and prevents any attempt to alter
data or commit repudiation. This addresses issues of trust in the complex process,
bringing in transparency, without need for oversight. This technology can solve
problems in military logistics related to networking, data storage, system
maintenance, traceability and quality control during packaging, loading and
unloading, transportation, and disassembly.

An example of the use of blockchain technology in military are automatic


execution of smart contracts, secure storage of sensitive files, and reduction of
errors and interruptions during defence contract execution, was demonstrated in
April 2016, by the US Department of Defense and its NATO allies.

Increased transparency
Unified and more
and granularity on
accurate price and
integrated margin by
demand forecasts.
sale.

Marketing
On time, and Sales Full data
undamaged integration
deliveries. with suppliers.

Logistics
and AI for
Distribution Transparency Procurement
and DM

Dynamic optimization of routing, Optimization of raw


freight contracting, vessel materials recipes based
sharing, reducing costs and Planning on forecasted prices
environmental impact.

Risk adjusted Full transparency on


end-to-end execution through
margin end-to-end digital
optimisation. control tower.

Figure 6: AI for Transparency and DM

144
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

Quality Control and Customer Relations

Quality control would be enhanced and perfected by AI. To cite an


example, companies are increasingly using AI-Chatbots for consumer service.
Similarly, the services echelons could benefit from such measures and reduce
manpower. Chatbots could give full information to customer units on the location,
status of their deliveries and also address grievances for fast resolution. These
could be used to elicit information of logistics stocks and processes from higher
HQ. Utility in technical support can also be well visualised.

It is reasonable to expect that AI powered customer experience could be


the norm in the days to come, and it is easy to perceive the teeth to tail ratio
potential that it offers. Voice agent (a speaker or a software tool able to
communicate) offers a voice-based service in tracking ordered parcels and
providing with the information about delivery times, locations, and other
information. Similar system with due security measures, could offer full visibility of
logistics support rendered, to the consumer units.

AI in Logistics Risk Management Strategies

The technologies that underpin SC 4.0 include blockchain, smart contracts,


applications of AI, Cyber Physical Systems (CPS), IoT etc. CPS are systems that
involve the integration of the computing and physical world through cyber-enabled
control mechanisms enabling ability to interact with and expand capabilities of
physical world through computation, communications and control. Any processes
that operate on an internet platform are of course, vulnerable to cyber threats,
which will also target all these technologies as well as AI itself. But the sweet irony
is that AI itself could be used to tackle these threats effectively. The importance of
Decision Support Systems (DSS) for Supply Chain Risk Management (SCRM) is
pertinent, with different operations research techniques and methodologies for
decision making for managing risks, focusing on multiple-criteria decision analysis
methods and mathematical programming. AI techniques can be applied in the
SCRM domain to analyse data and make decisions regarding possible risks.

It is expected that logistics and supply chains would be more lucrative


targets to the enemy than even the fighting forces, as it would have the effect of
paralyzing a force without the requisite support. AI could play an important role to
predict, detect, prevent, safeguard and respond to offensive cyber-attacks. Risk
management strategies for SC 4.0 (risk identification framework, its assessment,
decision/ selection of corrective action, and an evaluation phase) will assume due
importance. Crisis management through logistics flow optimisation is also called
for.

145
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

The Teeth to Tail Ratio (T3R) Windfall

AI will help reshape T3R freeing combatants for critical areas like Intelligence,
Surveillance and Reconnaissance (ISR). An appropriate aspiration for the
restructuring of our forces is to ensure that the „teeth‟ as it were, does not get
overwhelmed by the „tail‟. AI will radically transform so many logistics processes
that it would be possible to significantly cut down logistics manpower in the
foreseeable future. Such optimised manpower could either be used for critical new
raisings or could lead to force cuts itself bringing huge savings to the exchequer.

The Merger of the Operations and Logistics Verticals

The very efficiency in processes through AI would reduce human


interventions in many routine as well as DM fields. It would significantly reduce
complexities of staff work and throw up the possibility of an eventual merger of the
Operations (G) and Logistics (Q) verticals, having far-reaching implications in HR
management in the Armed Forces. DM would be much faster with a single GQ
branch manned commonly. Headquarters would be leaner at all levels. Intake can
be more versatile as officers will able to perform wider functions without the need
to be constrained through their Arms and Services or even between the three
Services. This may offer a common cadre progression model integrating the
operation logistics rather than purely logistics.

Training

Training would be an important facet for AI penetration in the Armed


Forces. The training itself would have two distinct facets, as outlined in Figure 7.

Training

Training on Training
AI through AI

Using
Robots

Using Learning
Management Systems

Using Immersive technologies


like AR, VR, IVR etc

Figure 7: Facets of Training


146
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

Rather than awaiting the absorption of AI technologies in the IA, we must


encourage up-skilling of all ranks to understand AI and promote its proliferation in
our military. This can be undertaken through study leave, detailment for AI courses
at IISc and the IITs (B-Tech, M-Tech, Doctoral and even Post-Doctoral studies).
Again, this should not be limited to officers from the technical Arms/Services but
open to all, to enable each Arm or Service to develop a pool of AI qualified officers
who can then lay down or formulate the agenda for the absorption of AI within the
particular vertical. JCO/NCO leadership courses can also have a primer on AI
technologies to give an overview of the coming revolution so that they are mentally
and academically get attuned to these changes.

One word of caution is to avoid attempting to develop in-house


capabilities for AI as it must be understood that this is a very specialized and
niche field and training should be limited to understanding potential and empower
planning to assimilate. Here, „letting the experts do their thing‟ without
transgressing own core competencies, will be both prudent and cost-effective.

The Roadmap

With regard to the logistics echelons, it will be appropriated to first adopt


complete jointmanship in logistics, followed by development of common Enterprise
Resource Planning (ERP) protocols, which will pave the way for faster and easier
adoption of AI across the three services in logistics as opposed to a incremental
approach.

Summary

Every few years, a new revolution comes along with potential to change
the paradigms of warfighting and AI is surely one such development. But the
success of any country‟s military is largely dependent on ability to absorb,
implement and exploit such technologies as opposed to mere intellectual
discussions on the same ad nauseam. The need of the hour is to jettison
anachronistic thinking and be early adapters of emergent technologies like AI.

Lessons of the COVID pandemic as it impacted logistics support worldwide


and also to a certain extent in the Armed Forces, need to be learnt and borne in
mind for the future. It is likely that such challenges in the future would be aptly
handled through AI-based logistics systems.

At government level too, there is the need for adequate R&D push, funding
and the nurturing of educational and scientific ecosystems that empower the
growth of AI and similar technologies. Within the Armed Forces, initiatives like
setting up the Army Design Bureau and the Army Technology Hubs (under aegis
of ARTRAC) are steps in the right direction, as these agencies act as effective
interfaces with industry and academia and facilitate the assimilation of cutting-
edge technology in the Armed Forces.

147
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

Endnotes

1. R. Fedasiuk, J. Melot, and B. Murphy, “Harnessed Lightning How the Chinese Military is
Adopting Artifical Intelligence,” Centre for Security and Emerging Technology (CSET), no. October,
2021.

2. C. Dilmegani, “Top 15 Use Cases, Applications & Examples of AI in Logistics,” AI Multiple,


2022. https://research.aimultiple.com/logistics-ai/.

3. D. W. Cearley and B. Burke, “Gartner: Top 10 Strategic Technology Trends for 2019,”
Gartner Research, no. October 15, 2018.

4. McKinsey and Company, “Supply Chain 4.0 – the next-generation digital supply chain,”
2016.

5. E. Krmac, “With Artificial Intelligence towards Intelligent Logistics and Supply Chains : the
state of the,” ICEST, no. June, pp. 198–207, 2019.

6. DHL Trend Research, “Logistics Trend Radar 5th Edition,” 2020. [Online]. Available:
https://www.dhl.com/global-en/home/insights-and-innovation/insights/logistics-trend-radar.html.

7. D. Khasis, “Four Ways AI Is Impacting Logistics and Supply Chain Management,” Supply
Chain Brain, 2019. https://www.supplychainbrain.com/blogs/1-think-tank/post/30045-four-ways-ai-is-
impacting-logistics-and-supply-chain-management.

8. E. B. Kania, “Chinese Military Innovation in Artificial Intelligence,” 2019.

9. Y. Zhu, X. Zhang, Z. Y. Ju, and C. C. Wang, “A Study of Blockchain Technology


Development and Military Application Prospects,” Journal of Physics: Conference Series, vol. 1507,
no. 5, 2020, doi: 10.1088/1742-6596/1507/5/052018.

10. D. Ge, G. Li, Z. Wei, and X. Wen, “Exploration of Key Technologies of Smart Logistics
Based on Big Data,” Advances in Social Science, Education and Humanities Research (ASSEHR),
vol. 184, no. Icesem, pp. 1256–1259, 2018, doi: 10.2991/icesem-18.2018.293.

11. T. Sobb, B. Turnbull, and N. Moustafa, “Supply chain 4.0: A survey of cyber security
challenges, solutions and future directions,” Electronics (Switzerland), vol. 9, no. 11, pp. 1–31, 2020,
doi: 10.3390/electronics 9111864.

12. P. Svenson, C. Martenson, H. Sidenbladh, and M. Malm, “Swarm Intelligence for logistics :
Background,” 2004.

13. D. Brahmabhtt, “How Artificial Intelligence Benefits Logistics and Supply Chain
Management,” Becoming Human: AI Magazine, 2021. https://becominghuman.ai/how-artificial-
intelligence-benefits-logistics-and-supply-chain-management-c32b43845af1.

14. L. Kota, “Artificial intelligence in logistics Applications and Algorithms,” Advanced Logistic
Systems, vol. 12, no. 1, pp. 47–60, 2018, doi: 10.1007/s13218-010-0022-6.

15. R. Siddharth, “Bringing the „Meta‟ change in supply chain through VR,” Logistics Insider,
2022. https://www.logisticsinsider.in/bringing-the-meta-change-in-supply-chain-through-vr/ (accessed
Jan. 30, 2022).

148
PINNACLE 2022
Artificial Intelligence and the Future of Indian Military Logistics

16. Dr. A. P. Pandian, “Artificial Intelligence Application in Smart Warehousing Environment for
Automated Logistics,” Journal of Artificial Intelligence and Capsule Networks, vol. 1, no. 2, pp. 63–72,
2019, doi: 10.36548/ jaicn.2019.2.002.

17. K. Alicke, V. Dilda, S. Görner, L. Mori, P. Rebuffel, and R. Samek, “Succeeding in the AI
supply-Chain Revolution,” 2021.

18. A. Kendall, A. Das, B. Nagy, and A. Ghosh, “Blockchain Data Management Benefits by
Increasing Confidence in Datasets Supporting Artificial Intelligence (AI) and Analytical Tools using
Supply Chain Examples,” in Eighteenth Annual Acquisition Research Symposium, 2021, no. May 11-
13, pp. 209–230. doi: 10.1080/19342039.2020.1822712.

19. Z. Davis, “Artificial Intelligence on the Battlefield,” PRISM, vol. 8, no. 2, pp. 114–131, 2019.

20. G. Baryannis, S. Dani, S. Validi, and G. Antoniou, “Decision Support Systems and Artificial
Intelligence in Supply Chain Risk Management,” Springer Series in Supply Chain Management, vol.
7, pp. 53–71, 2019, doi: 10.1007/978-3-030-03813-7_4.

21. A. Kaddoussi, N. Zoghlami, H. Zgaya, S. Hammadi, and F. Bretaudeau, “Disruption


management optimization for military logistics,” IFIP Advances in Information and Communication
Technology, vol. 364 AICT, no. PART 2, pp. 61–66, 2011, doi: 10.1007/978-3-642-23960-1_8.

22. I. Chakir, M. El Khaili, and M. Mestari, “Logistics flow optimization for advanced
management of the crisis situation,” Procedia Computer Science, vol. 175, pp. 419–426, 2020, doi:
10.1016/j.procs.2020.07.059.

23. P. K. Mallick, “Artificial Intelligence in Armed Forces: An Analysis,” CLAWS Journal, no.
Winter, pp. 63–79, 2018.

24. S. Modgil, R. Singh, and C. Hannibal, “Artificial Intelligence for Supply Chain Resilience:
Learning from Covid-19,” The International Journal of Logistics Management, 2021.

149
PINNACLE 2022
LEGAL AND ETHICAL ASPECTS OF ARTIFICIAL
INTELLIGENCE IN FUTURE WARS
- Colonel Surendra Tanwar

Introduction

As AI technology rapidly evolves, governments are increasingly exploring


its potential military applications. While some argue that AI could provide strategic
advantages on the battlefield, others raise concerns about the legal and ethical
implications of its use in warfare. The legal status of military AI is currently
unsettled, but there are growing calls for regulation in this area.

This article aims to explore the legal and ethical aspects of using AI in
warfare. It will first discuss how international law applies to AI-enabled weapons
systems. Next, it will consider the arguments for and against of using AI in combat
operations. Finally, actions taken by various nations to ensure that deployment of
AI-enabled weapons systems complies with international law and respects human
rights principles will be discussed.

Use of Artificial Intelligence in Military

The use of AI in the military has increased dramatically in recent years. AI


is being used in logistics, cybersecurity, simulation and wargaming, planning, and
surveillance.

However, as AI technology advances, its potential uses in the military will


also continue to grow. Some people argue that AI should not be used in the
military at all because of ethical and legal concerns. Others believe that AI can be
a valuable tool if used responsibly. While discussing the ethical and legal issues
associated with using AI in the military, the paper will focus on these issues arising
mostly out of the use of Lethal Autonomous Weapons Systems (LAWS).

International Humanitarian Law (IHL)

One key concern in use of Artificial Intelligence in warfare is whether or not


AI can be made to comply with IHL also known as the Law of Armed Conflict
(LOAC). IHL is a set of laws that seek to limit the effects of war on civilians, and it
includes provisions such as the prohibition on targeting civilians, the prohibition on
using weapons that cause unnecessary suffering, and the requirement that
combatants distinguish themselves from civilians. These principles can also be
termed as Precaution, Proportionality and Distinction.1

1
Morgan, F. E., Boudreaux, B., Lohn, A. J., Ashby, M., Curriden, C., Klima, K., & Grossman, D.
(2020). Military Applications of Artificial Intelligence: Ethical Concerns in an Uncertain World. RAND
Corporation.

PINNACLE 2022 150


Legal and Ethical Aspects of Artificial Intelligence in Future Wars

Can AI be programmed to comply with these provisions? Some experts


believe that it can. For example, Professor Toby Walsh from UNSW Sydney has
said: "There's no reason why AI could not operate within all sorts of legal
constraints like human soldiers do." However, others are more sceptical. Dr Stuart
Russell from UC Berkeley has said: "It's very hard to see how you would get an AI
system to reliably obey international law."

Critics of fully autonomous weapon systems argue that these weapons


would not be able to comply with LOAC’s principle of distinction, which requires
that combatants be able to distinguish between civilians and military targets. They
maintain that these weapons are not able to make the necessary ethical decisions
on their own, and could easily result in civilian casualties. Others argue that such
concerns are overblown, as these systems would be under human control at all
times and would only be used in situations where civilians were not likely to be
harmed. Distinguishing combatants from non-combatants might be especially
challenging in the context of asymmetric conflict in urban settings, where
combatants do not always wear uniforms or other insignia. Especially in these
settings, only a human operator can comply with the principle of distinction.2

Critics of autonomous weapon systems have also argued that they cannot
satisfy the principle of proportionality, which requires that any attack be
proportional to the threat posed. For example, an autonomous weapon system
might be programmed to destroy a tank, even if that would also kill innocent
civilians nearby.

Two counter arguments against these are firstly, any weapon can be used
in a manner that violates these laws and secondly, AI based agents maybe more
capable of strongly implementing these norms as against the human soldier, who
is prone to emotional biases and prejudices.

Concerns for Human Rights

Governments must ensure that any military use of AI complies with


international human rights law. This includes ensuring that AI is not used in a way
that causes unnecessary harm or death, discriminates against people based on
race, ethnicity, religion, or other factors, or violates people’s right to privacy. In
addition, governments must take steps to prevent misuse of AI by rogue actors.

The development and deployment of autonomous weapons systems raises


particular concerns in this regard. These are weapons systems that can select and
engage targets without human intervention. They could be easily abused by
dictatorships or terrorist groups seeking to target civilians indiscriminately. As

2
Morgan, F. E., Boudreaux, B., Lohn, A. J., Ashby, M., Curriden, C., Klima, K., & Grossman, D.
(2020). Military Applications of Artificial Intelligence: Ethical Concerns in an Uncertain World. RAND
Corporation.
151
PINNACLE 2022
Legal and Ethical Aspects of Artificial Intelligence in Future Wars

such, it is crucial that any development of autonomous weapons systems be


subject to rigorous scrutiny from both national authorities and international bodies
like the United Nations Convention on Certain Conventional Weapons (CCW).

Governments must also ensure transparency about how they are using AI
in warfare. They should disclose information about what kind of AI is being used;
how it is being used; who makes decisions about its use; and what safeguards are
in place to protect civilians from harm.3

Accountability and Moral Responsibility

Accountability and moral responsibility are essential when using artificial


intelligence in warfare4. Officials must be able to answer for their decisions, and
ensure that any AI-assisted actions comply with international law and morality.

The use of AI in warfare can raise difficult ethical questions. For example,
when a human is responsible for making a decision about whether or not to launch
a missile, they can be held accountable if something goes wrong. But what
happens if the decision is made by AI? Who is responsible then? And how do we
ensure that the AI’s actions are morally justified?

These are important questions that need to be considered before using AI


in warfare. Officials must be clear about who is accountable for any mistakes or
wrongdoing.

Interpretability

Another concern about the use of AI in warfare revolves around


interpretability of decisions. For example, if an AI system kills a civilian rather than
a combatant, is it because the AI was confused about who was a combatant and
who was not? Or did it deliberately choose to kill the civilian? Another issue is
interpretability. Decisions made by the algorithm are incomprehensible to people,
which resulted in the introduction of the so-called "right to explanation" meaning a
right to give an explanation for an output of the algorithm.5

Interpretability is important for two reasons. First, if we can't understand


why an AI made a particular decision, we can't be sure that it's behaving ethically.
Second, without interpretability, we won't be able to trust AI systems in life-or-

3
Eleanor Bird, Jasmin Fox-Skelly, Nicola Jenner, Ruth Larbey, Emma Weitkamp and Alan Winfield
(2020). The ethics of artificial intelligence: Issues and initiatives. European Parliament.
4
Morgan, F. E., Boudreaux, B., Lohn, A. J., Ashby, M., Curriden, C., Klima, K., & Grossman, D.
(2020). Military Applications of Artificial Intelligence: Ethical Concerns in an Uncertain World. RAND
Corporation.
5
De La Torre, L.F. The ‘Right to an Explanation’ under EU Data Protection Law. Available online:
https://medium.com/golden-data/what-rights-related-to-automated-decision-making-do-individuals-
have-under-eu-data-protection-law-76f70370fcd0 (accessed on 15 Mar 2022).
152
PINNACLE 2022
Legal and Ethical Aspects of Artificial Intelligence in Future Wars

death situations. If we can't trust that an AI system will make the right decision
every time, then it's not worth using it in warfare at all.

There are several ways to make AI more interpretable. One approach is to


use transparent algorithms, i.e., algorithms that are open source and easy for
humans to understand. Another approach is Machine Learning with Explanation
(MLE), which allows users to see how individual data points influence the
decisions made by an AI system. MLE has been shown to be effective at
explaining the behaviour of complex machine learning models. However, both
transparent algorithms and MLE suffer from scalability issues as they become
difficult or impossible to be used as datasets get larger.

A new technique called "local interpretation" may provide a solution to this


problem. Local interpretation breaks down large datasets into smaller chunks that
can be interpreted locally. This makes them easier for humans and AI to
understand without sacrificing accuracy or performance. Local interpretation could
help us build trustworthy AI that can make ethical decisions in high-stake
situations of active combat.

Trust and Reliability

Trust and reliability are two more important factors when using artificial
intelligence in warfare. One of the biggest concerns with using artificial intelligence
in warfare is trust. Can we trust that AI will do what we expect it to do? Will it follow
our orders faithfully, or will it go rogue? For this reason, we need to be sure that
we can trust AI before employing it in battle.

Another concern is reliability. Can we rely on AI to perform consistently


under pressure? In a chaotic battlefield situation, mistakes can cost lives. The last
thing we want is for AI systems to malfunction when they’re needed the most. We
need to make sure that they are reliable enough to count on when things get
tough.

Accuracy

AI has already been used in warfare in limited ways, with drones being the
most commonly known example. But as AI continues to develop, there are
growing concerns about its accuracy and its potential to be hacked. For example,
what happens if a country's AI-controlled drone fleet is hacked by another country?
Or what if an AI system makes a mistake that results in civilian casualties? These
are just some of the issues militaries need to consider when it comes to using AI in
warfare.

One concern is that hackers could take over an AI system and use it for
their own benefit. For example, they could send drones on missions they wouldn't
want the original country to know about or even cause accidents that results in
153
PINNACLE 2022
Legal and Ethical Aspects of Artificial Intelligence in Future Wars

civilian deaths. This possibility raises questions about who would be responsible if
something goes wrong: the person who designed the AI system or the hacker who
took control of it?

The authors M. Gregor; M. Hrubo; D. Nemec; J. Hrbek6 raised the


important issue of Lethal Autonomous Weapon Systems (LAWS) and strives to
answer the question of why AI systems should not have the right to decide about
killing people as part of warfare. The main problem that authors point to is the lack
of perfect, non-error-making AI systems. In the case of deciding about human life,
even the accuracy at a level of 99% is too small. So far, AI has not been 100%
accurate, so we need to ask ourselves whether or not we're willing to take the risk
of using AI combat situations.

Approach to Ethical and Responsible Artificial Intelligence in Defence

As AI technology advances, it is critical that nations develop ethical and


responsible approaches to AI to protect civilians, soldiers, and other parties
involved.

Nations have adopted different approaches in ethical aspects of AI military


applications. One approach is to develop codes of ethics for those working with AI.
This includes making sure that those creating AI algorithms are aware of the
potential implications of their work, and establish guidelines on ethical use of AI.
This could involve a code of ethics for its Defense Research Establishments,
which would include principles such as "respecting human autonomy" and
"safeguarding privacy".

Another approach is establishing an oversight body for regulating


application of AI in defence. This could involve setting standards or regulations on
data gathering practises and automation of weapons.

A third approach is increasing transparency around how AI is being used in


defence. This could involve sharing information about what data was used to train
an AI algorithm or publishing reports on how often particular AIs are being
deployed.

Each nation has taken a slightly different approach when it comes to using
Artificial Intelligence within Defence and Military Applications ethically and
responsibly, but they all share one common goal: protecting civilians from harm
while still using this new technology efficiently.

6
Šimák, V.; Gregor, M.; Hruboš, M.; Nemec, D.; Hrbček, J. Why Lethal autonomous weapon systems
are unacceptable. In Proceedings of the 2017 IEEE 15th International Symposium on Applied
Machine Intelligence and Informatics (SAMI), Herl’any, Slovakia, 26–28 January 2017; pp. 359–364.
154
PINNACLE 2022
Legal and Ethical Aspects of Artificial Intelligence in Future Wars

Conclusion

The ethical and legal issues of using artificial intelligence in warfare are
complex. On the one hand, AI could be used to make battlefield decisions more
quickly and effectively than humans can. This could lead to fewer casualties on
both sides of the conflict. On the other hand, there is a risk that AI could be used
to commit atrocities or engage in war crimes that would be difficult for human
beings to carry out.

There are also concerns about how AI might be used outside of actual
combat situations. For example, it is possible that AI will be used for surveillance
purposes, or as part of targeted assassination programs. There are also questions
about who will bear responsibility for any mistakes made by artificial intelligence
systems – the programmers, the military officials who deploy them, or someone
else entirely?

Ultimately, there are many unanswered questions about how artificial


intelligence will be used in warfare – and these questions need to answered before
any decisions about its use are made. Governments around the world should work
together to develop clear ethical and legal guidelines governing AI in warfare.

Bibliography

1. Military Applications of Artificial Intelligence Ethical Concerns in an Uncertain World by


RAND Corporation (2020).

2. Artificial Intelligence: An Introduction to the Legal, Policy and Ethical Issues James X.
Dempsey Berkeley Center for Law and Technology August 10, 2020.

3. The ethics of artificial intelligence: Issues and initiatives for European Parliament (2020).

4. Artificial Intelligence Applications in Military Systems and Their Influence on Sense of


Security of Citizens by Marta Bistron and Zbigniew Piotrowski in Electronics 2021.

5. Ethical Principles for Artificial Intelligence in National Defence by Taddeo, M., McNeish, D.,
Blanchard, A. et al. accessed from https://doi.org/10.1007/s13347-021-00482-3.

6. AI Principles: Recommendations on the Ethical Use of Artificial Intelligence by the


Department of Defence by Defense Innovation Board accessed from https://media.defence.gov/
2019/Oct/ 31/2002204459/-1/-1/0/DIB_AI_PRINCIPLES_SUPPORTING_ DOCUMENT.PDF.

7. The Ethics of War and Peace: An. Introduction, 2nd ed.; Routledge: London, UK, 2015 by
Frowe, H.

8. Responsible and Ethical Military AI by Zoe Stanley-Lockman for CSET.

9. Routledge Handbook of Ethics and War by Fritz Allhoff, Nicholas G Evans and Adam
Henschke.

155
PINNACLE 2022
BOOK REVIEW: ARTIFICIAL INTELLIGENCE AND THE
FUTURE OF POWER: 5 BATTLEGROUNDS BY
MR RAJIV MALHOTRA
- Lieutenant Colonel Manish Rawat

Basic Information

Title : AI and the Future of Power


– 5 Battlegrounds
Author : Rajiv Malhotra
Publisher : Rupa Publications India
Date of : 10 Jan 2021
Publication
Language : English
Hardcover : 488 pages
Cost : ₹575.00 (hardcover),
₹6114.00 (Paperback)

Introduction

The author has discussed modern technologies to include AI in his book.


He has imaginatively portrayed five battlegrounds which India should be cognizant
of if India wishes to attain global leadership in AI and harness it towards creating
wealth and welfare for the nation.

About the Author

Mr Rajiv Malhotra is an internationally acclaimed author and founder of


Infinity Foundation. An alumni of St. Stephens College Delhi & Syracuse
University, he has distinguished background in physics and computer science. His
notable works includes Breaking India (2011), Being Different (2011), Indra’s
Net (2014) and The Battle for Sanskrit (2016). His intellectual approach to infuse
the background in physics and computer science with his research on India’s
historical and future place in the world gives immeasurable credibility to this book.

Production Value

Artificial Intelligence and the Future of Power: 5 Battlegrounds is a 520-


page hardcover book. The book is written by Rajiv Malhotra and is published by
Rupa Publications India Pvt Ltd. The selections include the table of contents,

PINNACLE 2022 156


Book Review: Artificial Intelligence and the Future of Power: 5 Battlegrounds

introduction, acknowledgement, glossary, biographical information, and about the


author, printed on the cover. The cost of the book is ₹ 795/-.

Theme

General The term "artificial intelligence" is used by the author to refer to


a wide range of modern technologies such as machine learning, big data analytics,
data science, quantum computing, semiconductor technologies, nanotechnology,
neuromorphic computing, robotics, 5G, smart manufacturing, and so on. He
acknowledges these are distinct, but AI brings them together in ways that give
them greater collective power. The book maintains the conceptual framework and
the examples quoted by the author are well thought out and explained.

Language The language used by the authors is simple and easy to


understand. The arguments are supported by adequate examples or references,
and the correlation has been carried out in a simplistic manner. The book is written
in discussion as well as a questionnaire style of writing, which makes it easy to
comprehend and correlate with the changing scenario. The book is easy to
understand and readers can easily relate the arguments made by the author to the
intrusion of technology into our lives and society.

Content

Layout The book has been divided into two parts. In the first part, the
author discusses the Algorithm versus Being, and Part Two is attributed to
Battleground for India. Four battlegrounds are discussed in part one, which
constitutes' The Battle for Jobs, The Battle for World Domination, The Battle
for Agency, and The Battle for Self’. The final battle, ‘The Battle for India',
constitutes part two of the book, which comprises chapters six to ten.

Description of the Book The book begins with fundamental concepts in


the realms of Artificial Intelligence and machine learning including elucidations on
the notions of Big Data, Deep Learning, Neural Networks and so on. The author
analyses the involvement of machines and AI-based technology in various
spheres, which include healthcare, transportation, military, agriculture, education,
and financial services. He raises queries by saying that, on the one hand, AI is the
holy grail of technology which will solve problems across virtually every domain of
our lives, and yet, simultaneously, it is creating conflicts on a variety of fronts. The
vast canvas on which AI’s impact is being felt means that there is a need to
discuss its complex ramifications in a meaningful and accessible way. The book
has identified the pivotal role of AI in each of the battlegrounds, which has multiple
players with competing interests and high stakes.

157
PINNACLE 2022
Book Review: Artificial Intelligence and the Future of Power: 5 Battlegrounds

Part I: Algorithm vs Being

 Overview of AI Technologies Before launching into compelling


arguments for each battleground, the author provides an excellent
overview of AI technologies. He first discusses fundamental advances such
as machine learning, data analytics and AI gadgets, followed by AI
applications ranging from healthcare and agriculture to education, military,
and financial services. He has touched upon the controversial issues
surrounding AI, such as unfairness, lack of accountability, lack of
transparency and questionable ethics. The remarkable aspect of this
chapter is its detailed analysis in the realms of economics, geopolitics,
psychology, and metaphysics.

 Battle for Jobs Chapter two is dedicated to economic


dimensions, wherein the author has extensively used the term "Data
Capitalism," which considers big data as the new kind of capital asset. He
argues about the grave situation in the future concerning jobs and
employment. Automation and mechanization of various industrial
processes, coupled with the inability of the Indian education system to re-
educate its workers for the ‘jobs of the future,’ India could be possibly
recolonized by the future world powers. However, this colonisation could
happen based on the capabilities in the realms of AI and technology. The
possibility of exacerbating the buying power of consumers in developing
countries and eventually sabotaging economic growth is the anticipated
threat. The author has explained the pandemic effects and how they could
accelerate these trends and influence the equilibrium of world powers.

 The Battle for World Domination The role of AI in world


domination and the developments in the geopolitical arena have been
discussed. The author has explained "digital colonization" and described
how the technologically developed economies prey on the poor economies,
especially in Latin America and the African continent. The raw data mining
by the technologically developed economies of the poorer countries is to
evolve algorithms and become stronger, which is attributed to the highly
ignorant and corrupt leaders of the poor countries. The author has critically
analysed China specifically in AI development, wherein the entire nation-
building plan is based on AI and has made huge investments in the same
by utilising its infrastructure of roads, trains, seaports, and digital highways.
In the future, where AI defines global power, China will replace the USA as
the next global power.

 The Battle for Agency The author talks about AI technology's


influence on human motions. The author discusses devious conduct and
machine learning technology by global tech giants like Facebook, Google,
Microsoft, and Amazon for mapping human psychology. How do these
technologies carry out detailed profiling of people, societies, and
158
PINNACLE 2022
Book Review: Artificial Intelligence and the Future of Power: 5 Battlegrounds

communities as a whole and make them addicted to the instant


gratification, aesthetics, and artificial pleasures of these platforms? In his
analysis, the author talks about various concepts and theories that explain
the technological intrusion in the public space, the aestheticization of
politics and power, and the exploitation of these firms to achieve goals in a
practical sense.

 The Battle for Self This chapter deals with the domain of
metaphysics. The author explains how AI has successfully transformed
machines to behave and act intelligently, covering all biological processes.
He is critical to the digital industry's development, which has led to a battle
between algorithms and consciousness.

Part II: Battleground India

 Stress Testing India The author has discussed how India is


progressing towards AI domination and visualises the negative impact of
the stress factors such as high population, poor education standards,
chronic unemployment etc. He attributes all of it to the weakening of Indian
thought and creativity because of a prolonged repression and colonisation
period in the past. He argues that the Indian lead in IT innovations and is
becoming a market for American technology and Chinese hardware is a
wakeup call. The opportunities following the COVID-19 Pandemic and the
use of soft power by India to reclaim its lost position in the world order have
been critically examined.

 Technological Dependence The capability inferiority and how


short-sighted solutions have undermined India’s capabilities have been
critically examined. The author argues that despite having highly qualified
IT minds, India could not make relevant progress in the field of AI and the
cheap IT workforce employed by the tech giants has further undermined
India’s bid for intellectual property. The chapter has highlighted India’s
failed efforts to create jobs, which are attributed to lack of technology, a
well-educated workforce, population, and an end-to-end ecosystem.

 Digital Colonization This chapter provides in-depth insight about


digital colonization, which is attributed to ignorant intellect, academia, and
policymakers. The author describes how large-scale big data across
multiple sectors is being traded for high investments by foreign tech giants
and is not available to Indian IT start-ups to access for AI projects, which
invariably leads to dependence on US or European data bases, thus
leading to digital colonization. This unprecedented dependence has led to
an intoxicated Indian society, which has slavishly accepted the new digital
form of authority and tolerated its manipulation.

159
PINNACLE 2022
Book Review: Artificial Intelligence and the Future of Power: 5 Battlegrounds

 Psychological Hijacking The author has explained Vedic social


science and how years of colonisation and poor leadership have destroyed
the very fabric of national unity. The deep inferiority complex has
imprisoned Indian social science in western sociology. The author
propagates how the Vedic concept of Purushartha can be used to
decolonize the social sciences and explains how AI systems induce people
that have abandoned dharma and lost their moorings by using big data.

 How Robust is the Rashtra? The author has explained how anti-
nationals exploit India’s internal fault lines attributed to vote bank politics.
He further elaborates on how AI systems on social media platforms have
been developing algorithms to predict and exploit Indians. He has raised
the issue of AI proliferation and how these breaking forces are being
trained on AI to destroy the Indian social fabric. The looming Chinese
threat has been discussed in detail, as should be how India should be
responding to it. The author has suggested certain practical possibilities
using AI systems to unify the nation against the threats discussed
previously.

Verdict

This scholarly and well-researched book succeeded in communicating its


message through a gripping narrative. The book has ready accessibility to a
non-technical audience who have no familiarity with AI, which is a striking aspect
of the entire book. The book is a must read since it brings out some surprising
facts about the future of Indian civilization's vulnerabilities, social media intrusion in
our personal lives, and our relentless efforts to chase the capitalist glitter. The
authors' urge to identify the Indian societal fault lines, which to some extent, this
book does, is an eye opener and makes the readers censoriously evaluate the AI
developments in the future.

Recommendation

Overall, the book is informative and alarming, and that makes it must-read.
A small consolation is that the author says he is working on the next book, where
he provides solutions to the problems listed in this one. The author has mentioned
the United States of America and China as a benchmark for developments in AI
capabilities and articulated what lies ahead for India on its road to glory.
Substantiated by facts, reasoning, analogies, and witty observations, the book
serves as a call to action for public intellectuals to become more active and intend
to challenge the narrative that is being fed to the Indian diaspora about its
trajectory of becoming a technological superpower.

160
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

BOOK REVIEW: AI BY DESIGN: A PLAN FOR LIVING


WITH ARTIFICIAL INTELLIGENCE
- Colonel Jitender Kaushik

Basic Information

Title : AI by Design: A Plan for Living


with Artificial Intelligence
Author : Catriona Campbell
Publisher : Chapman and Hall/ CRC; 1st
edition
Date of : 05 May 2022
Publication
Language : English
Hardcover : 154 pages
Cost : ₹7694.00 (hardcover),
₹2181.20 (Kindle edition)

About the Author

Catriona Campbell is a renowned behavioural psychologist and a leader in


Human-Computer Interaction (HCI). Catriona co-founded London-based
experience design firm Seren in 2001. Since then, she has guided some of the
world’s best-known brands towards digital success, earning a place in the BIMA
Digital Hall of Fame thanks to two decades of outstanding contributions. Now
called EY-Seren, after being acquired by Ernst & Young in 2015, the company has
offices in several countries, assisting clients to create well-designed experiences.
Catriona earned her degrees from the Universities of Stirling and Glasgow and
also studied at the Sorbonne, Paris. Over the last 20 years, Catriona has been a
tireless advocate of technology design, driving R&D in the field at any opportunity.
She has also trained and employed some of the top talent ever to populate UX
and HCI, ensuring more people can experience and engage with digitised
products and services.

161
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

Description of Book

The idea of Artificial Intelligence (AI)—systems so advanced they can


mimic or outperform human cognition—first came to prominence in 1950, when
British computer scientist Alan Turing proposed an "imitation game" to assess
whether a computer could fool humans into thinking they were communicating with
another human. In 1956, John McCarthy, one of AI’s founding fathers, co-authored
a proposal that coined the term "artificial intelligence," referring to machines'
thinking for themselves. Since then, the rise of AI has been enabled by
exponentially faster and more powerful computers and large, complex data sets.
Applications such as machine learning, whereby a system identifies patterns in
large sets of data, have demonstrated the potential for AI to be practical and
profitable.

AI is one of the most rapidly advancing and controversial topics in scientific


research. The number of journal and conference papers referring to AI in the
Dimensions from Digital Science database increased by more than 600% between
2000 and 2019. The world is all geared up and making strides with AI, and AI has
become ubiquitous in fields such as medicine, education, and security. It is
important to take a pause and give conscious consideration to significant ethical
and technical challenges and come up with pre-emptive measures rather than
being reactive to overcome them upfront.

The book gives a thorough explanation of the challenges with AI and starts
to talk about how we may avoid the worst-case scenarios. This book will assist
readers in developing a critical perspective on technology and challenging the
myths they have heard about how horrible AI is. Instead, they could opt to seize
the initiative and develop better technologies through design and planning. The
book is also a critical and authoritative overview of the exciting future of AI that
serves as a wake-up call to humanity to work together to regulate the development
of AI for the good of humanity and, potentially, its preservation. This book
introduces the reader to Artificial Intelligence and its importance to our future.
Campbell uses behavioural psychology, explores technology, economics, and
real-life and historical examples to predict five future scenarios with AI.

Description of the Content

Today, artificial intelligence plays a role in billions of people’s lives.


Sometimes unnoticed but often with profound consequences, it transforms our
societies and challenges what it means to be human.

Consciously and unconsciously, we are using AI-driven software and apps


every day. However, AI is poorly understood and not just among the general
public. Even in the business world, where AI powers some of the most disruptive
technologies, few have a profound, firm grasp of AI, although AI has been around
far longer than the age of social media.
162
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

The book is divided into eight chapters. Chapter One of the book
introduces the readers to the stages of AI and Singularity. The first generation of
modern AI is called "Narrow-AI," such as Netflix recommendations or automated
chat bots. Using machine learning (ML) to teach itself by feeding lots of
information about one task, narrow-AI will continuously hone its accuracy. They
don’t have general intelligence like humans and can only work on that particular
task. More worryingly, the second stage, Artificial General Intelligence (AGI), will
exponentially be more potent than narrow-AI. AGI does not exist yet, but, helpfully,
it may tell us when it arrives. AGI has the potential to reshape the world. The
author has taken an example of the biggest killer of the world, i.e., hunger, and
posed a scenario where AGI would be able to do all the agriculture with AGI-driven
machinery and could alter the crop pattern depending on the weather, location,
etc.

The third and final stage of AI is Artificial Super Intelligence (ASI). There,
the author introduces the readers to the ‘Singularity’. It is the stage where AI
attains human levels of intelligence, when machines can think for themselves and
don’t need humans, when they have grown so powerful that they have exceeded
our human intelligence and have developed the capability to increase their
intelligence. This moment will be very critical for mankind. The author further
argues that the Singularity may not happen. Things could still go another way.
Humans may yet end up controlling AI like a highly trained house pet. But AI will
grow until it reaches a point where it has the capability to exceed human
capability. Does it matter when it happens, or does it matter if it happens?

Chapter Two of the book introduces the readers to the Future-Back


Methodology, in which instead of starting from today and looking forward, we use
the future as the starting point and move backwards in time. Three steps of the
Future-Back model are: align on the current state of AI evolution, look at the
options for the future and agree on one outcome; and create a roadmap to
manage AI.

In Chapter Three, "Should We Be Afraid of the Current State of AI?" the


author dwells upon the current state of AI in National Security and Defence,
Cyber-Warfare, AI Weapons, and Government sponsored AI Research and
Development. The author also discusses in detail the current state of AI in
business companies, technology companies, financial services, and AI-powered
healthcare. The misuse of AI by criminal gangs and terrorism has been adequately
covered. The biggest fear that AI will steal jobs has been adequately answered.
The author brings out that AI software will replace positions that don’t require
complex physical activity, so office/desk-bound roles will come under heavy attack.
Finance, HR, Legal, and other back-office functions in companies are already
increasingly outsourced and are not deemed core competencies by large
companies. Machine-produced goods and services will be cheap and plentiful for
163
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

all. Time alone will answer questions like: Are we heading toward a work-free
utopia, or will the coming robot age merely exaggerate the income inequality that’s
rampant across the globe?

Chapter Four describes the current state of AI Governance and


Regulations created at local, national, and global levels, and at the top of the
pyramid sits International Law. There are a handful of international law courts to
enforce international law against countries formally and, in some cases,
individuals. They are the UN International Court of Justice (ICJ), the UN Security
Council (UNSC), and the International Criminal Court (ICC). These courts are very
precise in their remit. There are some very thoughtful guidelines on AI from
global/regional organisations, although they are not enshrined in law. The most
influential being the World Economic Forum (WEF), Organisation for Economic
Co-operation and Development (OECD) and EU White paper on AI. They are well
written, reasonably detailed, and some countries and companies use them to
develop their AI laws and internal policies.

Chapter five talks about the current state of AI ethics. In November 2021,
the 193 Member States at UNESCO’s General Conference adopted the
Recommendation on the Ethics of Artificial Intelligence, the very first global
standard-setting instrument on the subject1. It will not only protect but also promote
human rights and human dignity, and will be an ethical guiding compass and a
global normative bedrock, allowing us to build strong respect for the rule of law in
the digital world. The author brings out that big companies like Microsoft, Google,
and Deep-Mind claim their teams are working on ethics to ensure that tech is built
and used responsibly, and that AI can benefit society without reinforcing bias or
unfairness. The author has brought out eight challenges that need to be answered.
A few of them, like if an AI arms race exists and left unchecked, the criminal
hacking which is already disrupting business and threatening to interfere with
politics; a handful of mega-companies will use their mastery of AI and data to
dominate their industries and, with reduced labour costs, attain unheard of profits
and contribute fewer tax receipts.

Chapter Six discusses the possibilities for our future with AI with the
readers. Five plausible scenarios have been painted to draw on diverse expertise,
including academics, businesspeople, politicians, and futurologists. Firstly, could
AI be stopped at this stage? which is not possible as AI is already pervasive in our
society. We are already ankledeep in narrow-AI technology challenges – deepfake
technology, AI hacking, and military AI weapons. Secondly, several movies have
initiated the thought that we are living in a simulation. This simulation theory itself
has a reason why there is no evidence of being in a simulation because any
evidence could be simulated. Thirdly, if we want to avoid fixing AI, then there’s
another big solution being bandied about right now: escaping Earth. Big

1
Ethicsofartificialintelligence-UNESCO,https://www.unesco.org/en/artificial-intelligence/
recommendation-ethic.
164
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

technologists like Elon Musk, with his Mars Spaceflight Architecture, Star ship,
hope that by 2030 we’ll set off to make this a reality, becoming a "multi-planetary
species".

Fourthly, the separation hypothesis, where a future AGI replaces a human


government, means opposition groups would likely emerge who want the power to
remain in human hands. Some groups would call for more human control before
this point, as daily life becomes more controlled by AI. An independent country or
state could then be founded, populated by humans who limit themselves to
narrow-AI applications or indeed none at all, and this separation hypothesis is
entirely possible. Fifthly, merging technology with the brain, which is part of
Neurotechnology, and the sub-field of Brain–Computer Interface (BCI), is relatively
new. The most famous BCI company, Elon Musk’s Neuralink, is working in this
direction.

Chapter Seven creates a roadmap—a plan for living with AI. In this
chapter, the author brings out the reality of AI incrementally creeping up on us until
we can’t live without it. Every device will be smart. So, we need to ensure that the
choices we make today move us in the right direction. We must plan for the
journey to arrive on time. The Singularity won’t be a surprise to us. There is an
inevitability to our merging with technology. Once we accept this, we need to plan
for the future. Short term guidelines (2022-2025), ethics, regulations, and policies
for AI governance, one agreed framework, and improving policymakers’ technical
knowledge can be a few of the steps towards embracing AI in a smoother manner.
The author has brought out one very valid point about the education of AI
engineers. This is very crucial if we’re to ensure safe, inclusive, and fair AI
systems. AI engineers understand the ethical elements they’ve coded into the
systems before going live. As bias is subjective, this is problematic, but there are
ways of educating people to provide clarity and instruction. Medium-term
milestones (2025-2028) like police AI to prevent misuse and protect people and
creating a Non-Proliferation of Nuclear Weapons style treaty for laws; medium-to
long-term milestones (2025-2030) like educating the world to understand and to
work alongside AI; reskilling the existing workforce; and managing the economic
consequences of AI have been amply batted.

Chapter Eight gives the reader a new hope for the AI. Giving some cause
for optimism, the author says, if we effectively manage the challenges of AI, there
are great benefits for humanity”. The author discus the role of purpose in life and
for individuals, work is probably the single most crucial factor for how people get
purpose on a day-to-day basis. Work is vital for many people to derive their sense
of purpose. But when your life’s purpose comes from work, when that work goes
away, it can be devastating. Here the author refers if people lose their jobs due to
AI. If Artificial Intelligence (AI) reduces or removes the purpose gained from work,
how do we replace it? But there is ray of hope and optimism as outside work,
people’s purpose focuses on family, friends, spirituality, hobbies and sport.
Parents spending more time with their children should improve relationships, boost
165
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

confidence and ultimately, create more opportunities – meaning that their talents
would flourish. People can look after their elderly parents. The author concludes
with the ray of hope stating that the Singularity could be next golden age for
humanity, where we take the benefits of AI and use it to design a better world and
by using the future-back approach, we design for a future that we want.

Recommendations

A nicely articulated book is worth reading for everyone who wants to


understand the current state of developments in AI and the impact it will have in
the future on human lives. The book clears the cobwebs in the mind, such as how
AI will affect jobs, whether AI will surpass human intelligence, and whether AI can
evolve to the point where humans have little or no control. This book is
recommended for students, common people who want to understand the tenets of
AI, working professionals in the field of AI, and academia who have to dwell right
away on the thinking of ethics and laws concerning AI.

Critique

Though a well-researched book which covers the current state of AI and its
ethics, options for the future, and a roadmap for living with AI, it does not give a
clear answer to how the fear of losing jobs due to AI will develop more wealth for a
common man whose job has been affected by AI. Although people will have more
time for themselves, their families, and to pursue their hobbies, how will they earn
their living after losing their job? Optimism for the future of using AI is clearly
visible in the last chapter, but at certain points it lacks backing from practical
aspects of human life and behaviour.

Conclusion

The book is a pleasant read and gives an insight and perspective into AI
encompassing the current state of AI, ethics and regulations, options for the future,
and a roadmap which may lead us to smoothly embrace AI in our lives. We are
currently at a tipping point where we can change and design the future before it
overtakes us-for good or for ill. The ultimate level of design sits with government
and regulatory bodies, and they have the power to make or break AI. We do know
that AI will, at some point, dominate every aspect of human life. Every country,
industry, and individual will be impacted by Artificial General Intelligence (AGI).
The world will never be the same. If we choose to grow with AI, we must decide
how to select the best possible future for humanity.

166
PINNACLE 2022
Book Review: AI by Design: A Plan For Living With Artificial Intelligence

PINNACLE THEMES SINCE INCEPTION

Issue Themes
Mar 01 Cost Effective Total Quality Training for the 21st Century
Soldier
Mar 02 International Terrorism
Sep 02 Revolution in Military Affairs in the Armed Forces
Mar 03 Involvement of the Non-uniformed Citizenry in the Affairs of the
Armed Forces
Oct 03 Involvement of the Armed Forces in the Long Term National
Development Projects
Apr 04 Specialization: Need of the Hour
Oct 04 Border Management - A Model
Apr 05 Doctrinal Approach to Sub Conventional Warfare
Oct 05 Synergy in Jointmanship
Apr 06 Transformation Doctrine for Indian Army
Jan 07 Leadership Challenges in The 21st Century
Oct 07 Dynamics of Short Intense War: Conceptual and Doctrinal
Ramifications for the Army
Apr 08 Internal Security Paradigm Emerging Challenges
Jun 09 China Defence Modernisation and Its Implications
Jun 10 Indian Army’s Role in Nation Building
Oct 11 Emerging Technologies and their Impact on Future Warfare
Dec 12 Training Challenges for a Transforming Army
Sep 13 Back to Basics - Need of the Hour
Nov 14 Shaping Tomorrow’s Military Leaders
Oct 15 Empowerment of the Indian Soldier for Future Conflicts
Oct 16 Next Generation Warfare
Oct 17 Role of Armed Forces in Meeting India’s Regional Aspirations
Oct 18 A Growing China – Implications for Asia
Oct 19 Manifestation of Grey Zone Warfare and Limited Wars Under
the Conditions of Informationisation in Indian Context and
Capability Development to Counter the Same
Oct 20 Utilisation of Space Capability by IA in Future Wars
Oct 21 Employment of Offensive Cyber as a Multiplication Tool for
Defensive and Offensive Operations by Indian Army
Oct 22 Artificial Intelligence and the Future of Warfare: Roadmap for
the Indian Army

167
PINNACLE 2022
CHANGING CHARACTER OF WAR
Introduction

General The strategic environment in the 21st Century is characterised by


two trends. Firstly, the increased instability in the international system which gives
rise to conflicts and secondly, the unprecedented use of technology and its impact
on economics and politics which influences the conduct of war. These two trends
have created an environment where threats are both diffuse and uncertain and
where conflict is inherent, yet unpredictable. In the field of warfare, both continuity
and change; evolutionary and revolutionary coexist, each shaping the other.

Defining Warfare In today’s environment, war can be explained as an


activity which involves the use of force or imminent threat of massive destruction,
has more than one belligerent and hence rests on contention. Warfare also
assumes a degree of intensity and duration and finally, it has an aim defined in
political terms, but which can be defined in militarily terms as victory.

Theme 1: Evolving Character of War

Strategy for War The strategy of war as in other strategies need to


define ‘Ways’, ‘Means’ and ‘Ends’. There can be no doubt that the ‘End State’ of
war is victory. What remains to be defined are the ‘Ways’ and the ‘Means’.
Understanding these two issues can enable us to understand the nature and
character of war and what differentiates the Nature and Character of War.

 Nature of War ‘Nature of War’ can be explained and understood


as the ‘Means’ to wage and prosecute war. Clausewitz defines war as an
act of force to compel our enemy to do our will. Securing the ‘End State’
requires the enemy to be rendered powerless; and that, in practice, is
victory. Intrinsic to the ‘Nature of War’ is friction, fog, chaos, violence and
danger. These characterisations of war have remained constant since
immemorial times and therefore ‘Nature of War’ is enduring and timeless.

 Character of War If strategy for war seeks to achieve victory as


the desired ‘End State’, employment of physical force as the ‘Means’ to
wage wars, then what is left to be established is the ‘Ways’ to employ
physical force to achieve victory. Describing the ‘Ways’ to achieve victory
through the use of force defines the ‘Character of War’.

Emerging Character of War Every era of human evolution represents a


distinct ‘Character of War’. The first weapons developed during the Bronze Age
and the Iron Age. Even as the ‘Age of Gunpowder’ transformed warfare, the
creation of ‘Mass Armies’ during the ‘Napoleonic Era’ further changed the scale in
which wars could be fought. This theme would seek to explore the following:-

PINNACLE 2022 168


Changing Character of War

 Evolution of Warfare Emergence of modern warfare can be


traced back to the 16th Century when science and technology was first
used. Evolution of warfare can be explained through historical overview of
distinct periods which also correspond to four techno regimes; Large
armies and mechanism, Industrial war and thermodynamics, cold war and
cybernetics and finally chaoplexity in the post-cold war era.

 Determinants of Character of War Nations prosecute wars with


a desired ‘End State’ which in political terms could be vague or nebulous.
In addition, the desired ‘End State’ would define the level of victory to be
achieved and limits to the use of force in terms of the geographic region
where decision is sought as well as quantum of force is to be employed.
This sub-theme would seek to explain how the desired politico-military ‘End
State’ developed from the political ‘End State’ impact the Character of War.

 Grey Zone Warfare Nations typically transcend from confrontation


to conflict. Nations are however, increasingly exploiting the space between
confrontation to conflict, defined as the ‘Grey Zone’ to secure their national
interests. This sub-theme would seek to explore reasons why nations are
attempting to exploit the ‘Grey Zone’ and how has activities in this space
impacted the Character of War.

 Large-Scale Full Spectrum Conflict Notwithstanding the


proclivity to engage in Non-Contact Warfare in the Grey Zone, large-scale
full spectrum conflict is considered the greatest threat to the nation. This
sub-theme would draw inputson the implications and relevance of large-
scale conventional operations.

Theme 2: Trends in Warfare

General War is influenced by interplay of specific characteristics and


strengths of the parties to the conflict in question and the political, economic,
technological, intellectual, social as well as the will of the leadership. The sub-
theme will attempt to explore the following:-

 Trends in National Security Nations resort to the use of force to


secure national interests to enhance national security. Often, national
security is an expression of the worldview of nations and their leaders,
which is constantly evolving. An appreciation of trends in national security
would provide insights into the concerns of nations and why they would
transcend from confrontation to conflict.

 Emerging Domains The lexicon of Multi Domain Warfare (MDO)


is increasingly finding acceptance amongst militaries and scholars alike.
Land, sea and air, the three traditional domains of warfare have now been
joined by cyber, space and Electronic Warfare (EW). This sub theme would
169
PINNACLE 2022
Changing Character of War

identify the ways in which emerging domains could potentially impact the
Character of War.

 Doctrinal Trends in Warfare Developments in the international


system and rate of pace of technological change will have a profound
impact on warfare. However, Artificial Intelligence, stealth, robotics, long
range precision targeting etc will finally have to be synergised on the
battlefield by way of a warfighting doctrine. This part will attempt to
visualise contours of the likely futuristic MDO Doctrine.

 Leadership and Change Management Every transformative


effort follows the framework of people, structures and processes. Change
is however, never easy and its success cannot be assumed. This sub-
theme will attempt to identify human resource challenges, structures and
methods to implement transformative changes thereby enhancing
battlefield effectiveness.

Theme 3: Capability Development

General Militaries adopt a long term perspective to develop capabilities.


The capability development is undertaken under the backdrop of finite resources,
technological constraints and uncertain trajectory of capability development by
potential adversaries. Besides, civil-military relations, structures, doctrines etc also
play a significant role in developing military capabilities. The sub-theme will
attempt to explore the following:-

 Indian Thoughts on National Security India is classified as a


status quoist and a non-revisionist nation which does not intend to use
military force to alter its political boundaries. Notwithstanding its pacifist
intentions, India has often been constrained to use force both within and
outside its boundaries. This sub-theme would try to establish how the India
views its compulsions to employ force as well as the role and evolution of
its security structures and institutions.

 Doctrinal Evolution in the Indian Context The Indian armed


forces have come a long way since independence. This part will examine
the evolution of the three services of the Indian Armed Forces in the past
75 years.

 Factors in Capability Development The Indian Armed Forces


are influenced and constrained by structural and environmental factors
which impact its capability development. This sub-theme will attempt to
identify the structural and environmental factors which impact capability
development of the Indian Armed Forces as well as make relevant
recommendations.

170
PINNACLE 2022
Changing Character of War

 Character of War in the Indian Context The Indian Armed


Forces are confronted with two traditional adversaries in a potentially
collusive scenario, even as it has to contend with numerous insurgencies.
Given the environmental realities, this part shall intend to explore the likely
character of future conflict in the Indian context.

Theme 4: Strategies for Future Conflict

General Since wars are not merely military endeavours but represent
whole of the nation approach, developing a unitary understanding of strategy for
war is fraught with risks. Strategy for war would involve synergising multiple
activities in the vertical levels with each level representing numerous actions in the
horizontal plane. The sub-theme will attempt to explore the following:-

 Purpose, Triggers and the Desired End State This part shall
attempt to answer the following:-

 At the political level what could be the purpose (as opposed


to triggers) of war for India, Pakistan and China?

 What could be the desired end state for India, Pakistan and
China at the political and military-strategic level?

 Contours of Escalatory Continuum This part shall seek to


address the following:-

 What activities could India, Pakistan and China undertake


as part of the DIME Paradigm to implement their respective
strategies for war?

 What are the likely contours of escalatory continuum from


confrontation to conflict?

 Strategy for Defeat Mechanism This sub-theme would intend to


gain insights into the following:-

 At the military strategic level how are India, Pakistan and


China likely to develop their respective defeat mechanisms to
achieve the political end state?

 What could be likely strategies for war at the military-


strategic as well as the operational level for India, Pakistan and
China?

171
PINNACLE 2022
GUIDELINES FOR THE AUTHORS
General In our endeavour to generate a debate on a particular issue,
over-riding preference will be given to theme articles during selection of articles for
publication. Only researched and well substantiated articles with end notes and
bibliography are likely to find favour for publication in Pinnacle. All ideas generated
must be covered in a comprehensive manner and taken to a logical conclusion, so
that a tangible road map is perceived by the reader.

Articles Articles in electronic format should preferably be written in Font


Arial 11 on MS Word and sent along with the softcopy on a CD. A synopsis of up
to 150 words is also required with each article alongwith bio-data in a paragraph,
email and mobile number of the author.

Layout of Articles The suggested layout of articles is as under:-

 Synopsis.

 Title page.

 Main body.

 Acknowledgements.

 Foot Notes/End Notes/References/ Bibliography.

 Appendices/Annexure/Tables (as applicable).

 Photographs/Illustrations/Drawings (as applicable).

Title Page The title page should also include the full name, address,
email-id, contact number and a brief bio-data of the author.

Main Body A minor staff duty as per service writing is not required.
Preferably no abbreviations/ acronyms should be used. However, should they be
used to avoid repetition of big terms/ names, the expanded form should be given
when first used, with abbreviations/ acronyms inside the brackets. The length of
the articles may vary between 2500-4000 words, which should appear on the
last page at the end of the main body. Articles may be written in passive voice/
third person, to the extent possible.

End Notes/References/Bibliography Detailed end notes and


references/ bibliography of books/periodicals/magazines/journals/newspapers/
published or unpublished material used in the articles, indicative of research
carried out by the author, must be given as per standard format followed in this
journal. This is very important.

PINNACLE 2022 172


Guidelines for the Authors

Appendices/Tables Tables should be numbered in Arabic numerals and


should have a title.
Certificates The following certificates, as per Paragraph 21 of SAO
3/S/2001/MI and equivalent orders for other Services, duly countersigned by the
author’s superior authority, should be attached to the articles:-

 No Objection Certificate (by the IO) I have perused the article/


book etc submitted by and I have no objection to it being published/
broadcast in ‘Pinnacle’ Journal duly edited by the Editorial Board where
required.

 Author’s Certificate It is certified that “I have not used any official


information and/ or material or any information obtained by me in my official
capacity in writing the article titled ‘............................’.

 Originality Certificate I hereby certify that the article titled


‘............’ has been written by me and is original in its contents. It has not
been published earlier in/ sent to any other journal/ publication. All relevant
references and bibliography have been given.

Book Reviews/Letters ‘Book Reviews’ of the latest books aligned to


the theme and ‘Letter to the Editor’ are also welcome. Critical review of articles
published in the earlier issues of PINNACLE are also solicited which will be
published under ‘the critique’ section.

Last Dates for Receipt of Articles The articles must reach the ‘The
Editor’, PINNACLE, latest by 30 Apr for the next Issue. Articles, along with bio-
data and synopsis may also be e-mailed to the editor at
theartracjournal@hotmail.com. Certificates duly signed will, however, need to be
forwarded through post only. Also, the articles alongwith requisite certificates may
be submitted online on Army Intranet ARTRAC webpage.
Editing The Editorial Board reserves the right to suitably modify the
articles, without reference to the authors, for reasons of clarity, style, accuracy and
space limitations.
Honorarium An honorarium of up to ₹4000/- will be paid to the
authors, for each article, depending upon its quality, content and length.
Honorarium will also be paid for ‘Book Reviews’ and ‘Critical Reviews’ @ ₹1000/-
and ‘Letters to the Editor’ @ ₹500/-. Authors must intimate the details of their
bankers to enable despatch of Demand Drafts.
Correspondence All correspondence to be addressed to The Editor,
‘Pinnacle’, Doctrine Branch (CS Sec), Headquarters Army Training
Command, Shimla-171003 (Himachal Pradesh), India.

173
PINNACLE 2022
FOR THE ATTENTION OF READERS

PINNACLE is published every year on 01st Oct i.e. ARTRAC Raising Day

All correspondence should be addressed to:-

The Editor
PINNACLE
Doctrine Branch (CS)
Headquarters Army Training Command
Shimla (HP)-171003

E-mail : theartracjournal@hotmail.com
The Journal is available on ARTRAC ADN website and indianarmy.nic.in
on Internet.
2022
VOL 21

Printed at : Headquarters Army Training Command

You might also like