WATER AND WASTE WATER RESOURCE

MANAGEMENT USING ARTIFICIAL INTELLIGENCE

TECHNIQUES
REPORT OF MINI PROJECT-I

Submitted in partial fulfillment of the requirements for the degree of

BACHELOR OF TECHNOLOGY

In

CIVIL ENGINEERING

By
MONU KUMAR (201CV133)

VIKAS YADAV (201CV161)

RAHUL KASHYAP (201CV129)

RINKU MEENA (201CV143)

Under the guidance of

Dr. SUBRAHMANAYA KUNDAPURA

DEPARTMENT OF WATER RESOURCES & OCEAN ENGINEERING

NATIONAL INSTITUE OF TECHNOLOGY KARNATAKA
 SURATHKAL, MANGALORE-575025
NOVEMBER 2022

DECLARATION

We hereby declare that the B.tech Mini Project – I Work Report entitled “ water and waste water
resource management using artificial intelligence techniques ” which is being submitted to the
National Institute of Technology Karnataka Surathkal , for the award of the degree of Bachelor of
Technology in Civil Engineering in the Department of Water Resources & Ocean Engineering , is an
authentic record of our genuine work done under the of Dr. Subrahmanya Kundapura . The material
contained in this report has not been submitted to any University or Institution for the award of any degree.

MONU KUMAR (201CV133)

VIKAS YADAV (201CV161)

RAHUL KASHYAP (201CV129)

RINKU MEENA (201CV143)

Place: NITK, Surathkal

Date:

2
 Certificate

This is to certify that this report entitled WATER AND WASTE WATER RESOURCE MANAGEMENT
USING ARTIFICIAL INTELLIGENCE being submitted by MONU KUMAR(201CV133) , VIKAS
YADAV (201CV161) , RAHUL KASHYAP(201CV129) and RINKU MEENA(201CV143) is accepted
as the record of work carried out by them as the part of a Mini project-I in partial fulfillment of the
requirements for the award of the degree of Bachelor of Technology in Civil Engineering of the
Department of water resource and ocean engineering, National Institute of Technology Karnataka,
Surathkal, Mangaluru.

Dr Subrahmanya kundapura Dr B R Jayalekshmi

Professor & Professor and head &
Mini project supervisor Chairman (DUGC)
Department of civil engineering Department of Civil
Engineering
National Institute of Technology National Institute of
technology
Karnataka, Surathkal Karnataka, Surathkal

3
 ACKNOWLEDGEMENT

It is with great reverence that we express our gratitude to our guide “Dr. Subrahmanya
Kundapura”, Assistant Professor in the Department of Water Resources & Ocean engineering, for his
precious guidance for his expert guidance, valuable suggestions, reference materials provided, and
continuous encouragement throughout our project work. The credit for the successful completion of this
project goes to his keen interest in timing guidance and valuable suggestions otherwise our endeavor
would have been futile.

4
 ABSTRACT

Artificial intelligence applications play a crucial role in improving environmental quality from all points
of view. Digital technologies have revolutionized our way of life as they are permeated to a capillary level
in our daily life. On the other hand, the data produced every second cannot be managed by a human mind
due to a certain physical and temporal impossibility, so artificial intelligence, and algorithms written by
men to perform human reasoning, can accomplish this arduous task. In this chapter, we will address the
potential of artificial intelligence to process important amounts of data and analyze existing relationships
also through a focus on the conservation capacity of one of the most precious resources: water. In the
current era, water is a significant resource for socio-economic growth and the protection of healthy
environments. Properly controlled water resources are considered a vital part of development, which
reduces poverty and equity. Nonlinear effects, stochastic dynamics, and hydraulic constraints are
challenging in ecological planning for sustainable water development. In this paper, Adaptive Intelligent
Dynamic Water Resource Planning (AIDWRP) has been proposed to sustain the urban areas' water
environment. Here, an adaptive intelligent approach is a subset of the Artificial Intelligence (AI) technique
in which environmental planning for sustainable water development has been modeled effectively.
Artificial intelligence modeling improves water efficiency by transforming information into a leaner
process, improving decision-making based on data-driven by combining numeric AI tools and human
intellectual skills. In AIDWRP, Markov Decision Process (MDP) discusses the dynamic water resource
management issue with annual use and released locational constraints that develop sensitivity-driven
methods to optimize several efficient environmental planning and management policies. Consequently,
there is a specific relief from the engagement of supply and demand for water resources, and substantial
improvements in local economic efficiency have been simulated with numerical outcomes

5
 TABLE OF CONTENTS

DECLARATION
CERTIFICATE
ACKNOWLEDGEMENT
ABSTRACT
TABLE OF CONTENTS

1 INTRODUCTION
1.1 GENERAL

2 HOW CAN WE MAKE OUR CITY WATER SECURE?

2.1 IDEAS FOR WATER SECURITY IN DEVELOPING

COUNTRIES

3 A HOW CAN AI HELP?

4 SAVING WATER WITH ARTIFICIAL INTELLIGENCE

4.1 COLLECT DATA EFFICIENTLY
4.2 DO SMART THINGS WITH COLLECTED DATA
4.3 WATER TREATMENT
4.4 DETECT LEAKAGE
4.5 WATER MANAGEMENT IN HOMES AND BUILDING
4.6 AVOID EXPENSIVE MAINTENANCE

6
5 THE EMERGENCE OF AI IN THE AGE OF SUSTAINABLE DEVELOPMENT
6 IMPLICATIONS FOR MANAGEMENT EDUCATION
7 THE PROMISE OF AI IN WATER TREATMENT
8 ARTIFICIAL INTELLIGENCE IN WATER TREATMENT AND MONITORING
8.1 AI METHODS AND SMART TECHNOLOGY

8.1.1 AI METHODS

8.1.2 SMART TECHNOLOGY -THE INTERNET OF THIMS (IOI) &SMART

SENSING TECHNOLOGY
8.2 APPLICATIONS IN WATER & WASTE WATER TREATMENT

8.2.1 CHLORINATION & DISINFECTION BY-PRODUCT MANAGEMENT

8.2.2 ABSORPTION PROCESS

8.2.3 MEMBRANE FILTRATION PROCESS

8.3 APPLICATION IN WATER QUALITY MANAGEMENT

8.3.1 WATER QUALITY MANAGEMENT

9 MANAGING WATER WASTAGE

9.1 AI & INTEGRATED MANAGEMENT OF THE WATER CYCLE

10 CHALLENGES & PROSPECTS

11 RESULTS AND DISCUSSION
11.1 THE PROSPECTIVES OF AI IN THE WATER SECTOR

12 TYPICAL DESIGN OF ISOLATED FOOTING

13.1 PROBLEM STATEMENT

13.2 DESIGN CALCULATION AND REINFORCEMENT DETAILS

13 CONCLUSIONS

7
 INTRODUCTION

1. General
Promoting a circular economy through a ‘City Water Balance Plan’, focusing on recycling and reuse of
treated sewage, rejuvenation of water bodies, and water conservation are some of the key goals of this
mission. Leveraging the latest global technologies concerning water is another focus area. While the
scheme is expected to renew interest in using digital technologies in the water domain, it is of extreme
importance to look at urban water holistically and integrate climate change-induced risks in proposed
city water plans
Artificial intelligence (AI) is rapidly opening up a new frontier in the fields of business, corporate
practices, and governmental policy. The intelligence of machines and robotics with deep learning
capabilities have created profound disrupting and enabling impacts on businesses, governments, and
society. They are also influencing the larger trends in global sustainability. As the AI revolution
transforms our world, it could herald a utopian future where humanity co-exists harmoniously with
machines, or portend a dystopian world filled with conflict, poverty, and suffering. More
immediately, would AI accelerate our progress on the United Nations (UN) Sustainable
Development Goals (SDGs) or bring us further down the path toward greater economic uncertainty
environmental collapse, and social upheaval? What are some of the implications for business
leadership and the education of future business leaders? This article aims to address these questions
by analyzing the impacts of AI in three case studies. It draws some preliminary inferences for
management education and the business of leading corporations during rapid technological and
social change. This study combines the perspectives of business strategy and public policy to
analyze the impacts of AI on sustainable development with a specific focus on the advancement of
the SDGs. It also draws some lessons on managerial learning and leadership development for global
sustainability.

8
2. How can we make our cities' water secure?

Ideas for water security in developing countries

2.1 | Retain water as a public good

Privatization has been promoted globally for cities, but evidence shows that it often fails and the public
sector is left to address the problems. Cochabamba in Bolivia is an iconic case: enormous public protests
led to the cancellation of a private contract for the water utility. If water is a private commodity, then
people are at the mercy of market logic. Instead, ensuring that the principle of the water right is adopted
will enable more comprehensive thinking that addresses the needs of the urban poor, as well as ecological
sustainability.

2.2 | Diversify solutions

The important message to convey to those working in the sector and urban planning is that one size does
not fit all. Urban planners often aspire to develop fully networked and centralized water supply systems,
but this might not be the right approach for every city – big or small – as infrastructure layout is influenced
by geographical, political, and social boundaries. It is a reflection of the city’s history.

2.3 | Collaborate

The planning ministers and water ministers in most countries I’ve been to don’t speak to each other, so
we need better integration with knowledge exchange and the creation of specific agencies. Adding water
management to the training of the next generation of urban planners is also a way of ensuring that they
start thinking more about waste and water issues.

2.4 | Recycle wastewater

Sewerage and sanitation – and how water and wastewater can be treated and reused – have to be part of
any discussion around water security in cities. Recycling greywater, treating wastewater (or using it for

9
other purposes, such as fish farming), and understanding the complete water system is essential to
achieving urban water security.

2.5 | Legalise water vendors

In Maputo, Mozambique, the government and the water utility have worked together with small-scale
private water providers towards expanding the utility’s network into the poorest areas of the city. These
private operators, once illegal, were given formal status and training to ensure they provide a good quality
service. This initiative was only possible because both sides recognized the potential of working together,
but it’s important to ensure that governments remain the key regulators of this engagement.

2.6 | Connect informal settlements to the grid

A third to half of the populations in global south megacities are not connected to the water grid. The urban
poor continues to pay much higher costs for water (often from water vendors) compared with wealthier
neighborhoods that are on the grid. Informal water systems need to be part of the conversation, and
affordable housing can enable these millions of people to become part of the city, have tenurial rights, and
be connected to the water grid.

2.7 | Promote water efficiency in agriculture

In some modeling work I’ve done with the University of Kassel, most of the urban water gap – the shortfall
between projected demand and what people have to access now – could be closed by the agricultural sector
getting more efficient with water use. The question is, what sets of incentives or policies can help make
that happen?

10
2.8 | Understand affordability

You can look at the equity issue from multiple dimensions but in terms of affordability, Singapore is a
unique example and has some interesting tariff structures and subsidies targeted at lower-income
households.

2.9 | Use smart technology

We are entering a time when cities, individuals, and companies can incorporate new technologies that will
ensure water security for the future. Cities can use low-cost sensors and software for optimizing water use
and detecting problems in systems, for example. We are seeing several interesting pilots and initiatives
happening around the world with technologies that will continue to get cheaper

2.10 | Create an open data platform for water

Sensors and modeling of quantity in water bodies are getting much better but there is still a big gap in
knowing the human infrastructure – which water withdrawals are happening, where, and when. I think the
world needs something like an Open Water Map, where we pull together all the geospatial data on water
withdrawals –particularly cross-watershed ones – into an easy, publicly-accessible form.
Robert McDonald

2.11 | Promote water justice

In the context of climate change and rapid global changes, we need to keep in mind what kinds of cities
we want. To ensure water justice, communities can work with states and municipalities to manage and
conserve water. The important thing is to ensure meaningful (not token) participation, take into account
gender, class, and other social differences, and embrace social equity in planning and policies.

11
3. How can AI help?

AI can understand complex interrelations and can build dynamic and evolving city water models.
Possibilities of trade-offs between water allocation (for city use) and environmental flows (for
maintaining habitats, groundwater replenishment, etc.), predicting climate change-induced disaster
risks and suggesting potential nature-based solutions could be possible outcomes of such models.
Big data sets on weather patterns, water usage across sectors, water and wastewater networks,
reservoir levels, groundwater availability, existing and proposed land use plans, traffic data, etc.,
could be fed into AI models to identify patterns, inform decision-making, and even suggest possible
solutions. This could lead to the development of a dynamic and evolving City Water Balance Plan.
For example, the AI-based plan could inform potential consumers (like industries, commercial, or
residential complexes) of different qualities and quantities of wastewater generated. This can lead
to more reuse, and in turn, lesser freshwater extraction. These AI features could make city water
plans interactive and impactful for all stakeholders and facilitate stakeholder interactions in areas
earlier unrecognized.

Artificial intelligence (AI) is rapidly opening up a new frontier in the fields of business, corporate
practices, and governmental policy. The intelligence of machines and robotics with deep learning
capabilities have created profound disrupting and enabling impacts on businesses, governments, and
society. They are also influencing the larger trends in global sustainability. As the AI revolution
transforms our world, it could herald a utopian future where humanity co-exists harmoniously with
machines, or portend a dystopian world
filled with conflict, poverty, and suffering. More immediately,
would AI accelerate our progress on the United Nations (UN) Sustainable Development Goals
(SDGs) or bring us further down the path toward greater economic uncertainty, environmental
collapse, and social upheaval? What are some of the implications for business leadership and the
education of future business leaders? This article aims to address these questions by analyzing the
impacts of AI in three case studies. It draws some preliminary inferences for management education
and the business of leading corporations during rapid technological and social change. This study
combines the perspectives of business strategy and public policy to analyze the impacts of AI on
sustainable development with a specific focus on the advancement of the SDGs. It also draws some
lessons on managerial learning and leadership development for global sustainability.
12
Previously consigned to futuristic societies created in the imagination of science fiction writers and
motion picture producers, artificial intelligence (AI) is now a fact of everyday life in our modern
high-tech societies. There are many definitions of AI and each of those definitions have been revised
over time. Presently, most definitions state that AI solves complex cognitive problems associated
with human intelligence, that AI helps as many people
as possible through smartphones or healthcare,
or even that AI recognizes problems and creates solutions for the benefit of technology, people, and
society. However, the core concept of AI has continuously been to create machines that were capable
of thinking like humans (Marr, 2018).

AI is now increasingly ubiquitous in business and industry. It has the potential to revolutionize the
way we discover, learn, live, communicate, and work. It has tremendous potential for the economy
and society (National Artificial Intelligence …, 2016). As we enter the Age of Sustainable
Development (Sachs, 2015), in which the 17 Sustainable Development Goals (SDGs) are defining
the development agenda for the nations of the world, AI is also rapidly opening up a new frontier in
the fields of business, corporate practices,and governmental policy. The intelligence of machines and
robotics with deep learning capabilities is already solving cognitive problems commonly associated
with human intelligence.

13
“It took approximately 200,000 years for humankind's intelligence to evolve from natural to
artificial, and 10 years to cut the ties with ‘earth’ to move to the ‘cloud’” (Garimella & Fingar, 2018,
p. 7). During the creation of AI, humankind has learned a lot about what it means to be human, how
human intellect is structured, and how humans learn and gain expertise. The idea that 10,000 h of
deliberate and structured practice with the right quality is essential for achieving expert status was
first discovered by Anders Ericsson, a Swedish psychologist, and researcher on the psychological
nature of expertise and human performance, and popularized by Canadian journalist Malcolm
Gladwell (Garimella & Fingar, 2018). AI is rapidly moving into this world of expertise step by step,
by replacing humans with higher predictive power, higher efficiency, and better results. However,
homo sapiens will not be replaced all at once, but gradually over time as we become more absorbed in
the systems of knowledge and as robotics become an essential part of our human body and existence
(Harari, 2017).

The growth of AI will give intellectual and financial advantages to some cities and countries, while
others will be left behind. The rapid expansion of AI is already outpacing the development and
deployment of legal and regulatory frameworks and the mechanisms that are designed to govern it
(Munoz & Naqvi, 2018). Most experts who develop these frameworks and mechanisms think in the
short timeframe of academic research and/or political cycles, where a very long time may be as little
as 20 years (Harari, 2017).

With technological advances (such as those in the realms of computer vision, robotics, and speech
recognition), scientists, business people, governmental officials, and policymakers are becoming
increasingly concerned that AI will replace human workers, automate warfare, and supersede the
intelligence of humans (Markoff, 2014). According to Stephen Hawking, physicist and cosmologist,
humans with their slow biological systems could not compete with intelligent machines and could
easily be replaced by them (Goralski & Górniak-Kocikowska, 2017, 2018; Goralski & O'Connor,
2018; Penn, 2017).
14
 Dr. Eric Horvitz, President of the Association for the Advancement of Artificial Intelligence (2007–
2009) and Technical Fellow, and Director of Microsoft Research Labs include research centers in
Washington, Massachusetts, New York, Canada, the United Kingdom, and India, together with his
wife Mary, decided to fund the “One Hundred Year Study of Artificial Intelligence” (Markoff, 2014).
Together with Stanford, Harvard, and Carnegie Mellon Universities, the University of British
Colombia, and other partner institutions,

this study of AI will investigate whether the veracity of the fears and anxieties about AI and its threat
to humans are valid (Markoff, 2014, para. 4–5). Horvitz remains unconvinced by warnings of our
foremost thinkers and opinion leaders that super-intelligent machines are poised to outstrip human
control and abilities. He believes that these technologies will have both positive and negative effects
on society (Markoff, 2014).

While people in advanced countries may fear job loss due to AI, in low-income countries people may
see AI as offering new opportunities to break the cycle of poverty (Lohr, 2018). “The growth of AI
is rapid, selective, and impromptu
…. No one will be able to escape the ominous rise of AI” (Munoz & Naqvi, 2018, p. 1). The
academic community will have an important role to play in preparing future generations of business
leaders and national and international policymakers. AI and its positive and negative impacts must
be taught now so that students will be cognizant of the world as it is currently and the world as it
will rapidly unfold.

4. Saving Water with Artificial Intelligence

15
Researchers predicted that water will be scarcer than oil by 2030 and demand is going to exceed supply
by 40%. Less than 1% of all water on earth is fit for human consumption. The question for us is how we
can continuously monitor water everywhere without locating a lot of manpower to it.

We live in a data-driven economy and AI-driven world where everything is powered by data and artificial
intelligence. There is a paradigm within AI known as deep learning that specially focuses on creating
models based on the human brain to achieve efficient operations. Artificial intelligence measures data
analyzes, learns, and deduces.

Measure → Learn → Infer

Following are the water problems that can be solved by artificial intelligence:

16
4.1- Collect data efficiently:

AI technology helps to collect lots of real-time data to properly manage the water industry. Researchers
are using internet-connected wireless sensors to collect data. These sensors can measure temperature,
pressure, water flow, soil moisture, and more. These are connected to the internet from where anyone can
stream data to the cloud. This ecosystem of internet-connected sensors is called IoT. These AI-powered
sensors can measure vital things and that’s how we can collect data very efficiently.

Aquasuiteis a Netherland-based smart water technology company, which applies AI techniques, such as
machine learning and predictive analysis, to monitor, analyze, and control water performance.

4.2- Do smart things with collected data:

Billions of sensors and meters generate Exabytes of data. Artificial intelligence helps in extracting
meaningful insights from the data and guides in altering behavior for effectively managing water. AI helps
researchers to:

Track trends and Reduce the risk of water shortage

Identify anomalies
Forecast demand and Predict events
17
 Avoid water loss
Maintain a stable water supply
Sewer management
Cut the water usage

4.3- Water treatment:

AI can guide researchers to monitor water networks and wastewater treatment. Companies are designing
AI-based models to monitor water quality, process wastewater, remove pathogens, and make water fit for
drinking.

Fair cop AIis a social enterprise that develops new water filtration technologies to provide clean drinking
water solutions at an affordable cost. By using artificial intelligence in integration with other technologies,
the company aims to create large water monitoring datasets for rapid identification of potentially water-
borne diseases using a smartphone application.

4.4- Detect leakage:

Gallons of water get lost every day around the world because of inefficiencies like leakages, wastage,
aging pipes, and many more reasons. AI empowers researchers with the ability to perform real-time
analysis of water flow with algorithms and identify water leakages due to any fault in water infrastructure.
It assists in taking real-time decisions.

Facts , a California-based startup, is using artificial intelligence to detect leakages in pipelines and valves
and helps to make better asset management and maintenance decisions.

4.5- Water management in homes and buildings:

Municipal corporations and societies are taking advantage of artificial intelligence to manage water supply
properly. AI collects water usage data in homes and buildings and helps in designing water- and energy-
saving programs.

18
AI can monitor water usage by utilizing flow sensors in residential buildings. Flow sensors can be
deployed in pipelines to measure the water flow data and start streaming data to the cloud. This helps in
designing efficient water utilization methods.

4.6- Avoid expensive maintenance:

AI is being leveraged by engineers and water management authorities to keep water equipment and
pipeline networks in the right order and provide cost-effective maintenance solutions.

Pipebots is a UK-based project that aims to develop AI-powered micro-robots to analyze the state of
underground pipelines and affordably repair them.

Similarly, IBM, a global technology company, has utilized AI and machine learning to develop an in-pipe
inspection prototype named Sewer Spy to find erosion and corrosion of pipelines.

5. The emergence of AI in the Age of Sustainable Development

Jeffrey Sachs, professor of health policy and management at Columbia University, postulated that
the world is entering a new Age of Sustainable Development, an epoch in which the nations of the
world must collaborate and contribute to addressing the most intractable problems of persistent
extreme poverty, social exclusion, economic injustice, poor governance, and environmental
degradation (Sachs, 2015). Sachs was also a senior advisor to the UN on the SDGs and the
Millennium Development Goals and
5.1 Artificial intelligence and the SDGs

Artificial intelligence has been incorporated in various forms into the SDGs first through
experimentation, and later in sustainable management and leadership programs. This paper will
discuss three of those projects:

AI and the Water Crisis: Smart water management systems powered by AI replicate the way that
humans learn in an ever-changing environment that maximizes decisions and investments in the
water management infrastructure (Hill, 2018).

19
 AI and Agriculture: PlantVillage, a research and

AI and the Water Crisis: the case of smart water management

“Water is essential for life. For thousands of years, human settlement and advancement have been
dictated by a reliable supply of clean, safe water. In the face of a fickle supply system, people
flourished, moved or perished” (Hill, 2018, para. 1). Water and sanitation are at the center of the SDG
resource triad. It touches on, and impacts, women's empowerment and gender equality, food and
agriculture, energy and climate, and infrastructure & technology.

6. Implications for management education

While AI can have powerful applications for advancing various Global Goals, what are the
implications for management education? Tomorrow's managers and business leaders need to be
prepared to understand, and intelligently embrace the opportunities and challenges presented by the
new waves of technologies as effective vehicles of growth. There are already very significant
disparities among countries in their readiness for the AI revolution, and hence their capacity to
capture the potential

A plethora of global and Indian case studies present examples where AI intervention has yielded
results, albeit in distinct phases of the water cycle. For example, AI-based flood management
systems have proven successful in pilot installations in Portugal and France. AI technologies have
been used to detect patterns and leaks in city water systems, thus reducing non-revenue water (NRW)
— like in a county in South Korea that reduced water leakage by about 20%. Indian start-ups are
also developing AI-based real-time monitoring systems for the management of treatment plants,
rainwater harvesting, etc., though mostly at building and neighborhood scales. On a city scale,
Chandigarh and Pune have started shifting to a smart meter-based system for water distribution.
India has witnessed tremendous losses in recent years due to climate change impacts. Targeted

20
missions such as AMRUT are opportune in bringing much-needed attention to our climate change-
induced water conundrums.

FIGURE 1

It provides an opportunity to spot early on the synergies between climate change, development, and
water-related outcomes to maximize cross-sectoral coordination and stakeholder inclusivity. This
necessitates better water-related data and observation systems to guide climate resilience and water
resources management planning. AI has the potential to understand and respond to the inherent
complexities in water cycles and the exacerbated impacts that climate change has on the same. While
piece-meal AI and technology interventions may garner specific results, an integrated and holistic
intervention alone can ensure city-level impacts and make cities' water secure.

Artificial intelligence (AI) is making its mark on the water industry. It is powering intelligent
operations using machine learning to optimize resource use and operational budgets for
organizations, as well as delivering truly intelligent built water systems. Provided below are 10 ways
that AI is changingthe water industry.

1. AI will drive a decade of technology investment in water and wastewater operations

21
Water and wastewater operations are investing in artificial intelligence (AI). Recent market research
is forecasting $6.3 Billion of investment in AI solutions by the year 2030. This investment is a part
of a growing trend for the water industry to “go digital” with smart infrastructure solutions.

2. AI will deliver significant OPEX savings in water and wastewater operations

When US utilities are spending ~$300 per annual customer on water and wastewater operations, the
potential for savings is significant. AI can save 20- 30% on operational expenditures (OPEX) by
reducing energy costs, optimizing chemical use for treatment, and enabling proactive asset
maintenance.

3. AI will predict emergency events and learn from them at an accelerated rate

Water main breaks are costly for utilities – in both financial and social capital. AI and machine
learning can “fingerprint” the data patterns that indicate a break event may be imminent and learn
from these patterns so that alerts become more accurate over time.

4. AI will provide sophisticated decision intelligence to support operators

Operators no longer need to analyze complex variables for important decision-making by

themselves. Whether it’s turning pumps on or off, determining chemical dosages, or deciding
when to maintain assets, AI

enables the Operator 2.0 – empowered by intelligent recommendations driven by machine

learning.

5. AI will optimize energy use for water and wastewater operations

According to the USEPA, energy consumption makes up 25-30% of total operation and maintenance
(O&M) costs. AI can optimize pump runtimes so that they are only using energy when they need to.
This is a quick cost- reduction win for many early adopters of AI.

6. AI will keep the water clean, cost-effectively

22
 Meeting effluent compliance standards is a requirement for many organizations – both public and
private. AI learns from the unique characteristics of your site to ensure that effluence standards are
met and that compliance fees are avoided.

7. AI will simplify data integrity

An explosion of available data for water operations managers has brought along a challenge in data
management. SCADA systems, CMMS, and even social media hold a wealth of knowledge to
improve operations. AI can take this heterogeneous data and process it so that it is clean, useful,
secure, and drives high-fidelity recommendations.

8. AI will retain institutional knowledge

How do you ensure that a veteran operator’s valuable knowledge is retained after they leave the
workforce? AI-powered dashboards will keep institutional knowledge documented and standardized.

9. AI will accelerate the move to value-based asset maintenance

Early adopters of AI are quickly leaving reactive asset maintenance behind. Time-based
maintenance is easy to manage but results in unnecessary uptime and deterioration. Let AI tell your
team what assets need to be serviced, and when.

AI will power truly smart water systems The journey to AI adoption empowers organizations
to pursue data-driven, intelligent management of water systems. The result is resilient, sustainable,
and cost-effective water management for years to come

23
7. THE PROMISE OF ARTIFICIAL INTELLIGENCE IN WATER
MANAGEMENT

Artificial intelligence (AI) has emerged as a powerful tool to resolve real-world problems and has
gained tremendous attention due to its applications in various fields. In recent years, AI techniques
have also been employed in water treatment and desalination to optimize the process and offer
practical solutions to water pollution and water scarcity. Applications of AI are also expected to
reduce the operational expenditures of the water treatment process by decreasing the cost and
optimizing chemicals usage. This review summarizes various AI techniques and their applications in
water treatment with a focus on the adsorption of pollutants. Numerous AI models have successfully
predicted the performance of different adsorbents for the removal of numerous pollutants from
water. This review also highlighted some challenges and research gaps concerning the applications
of AI in water treatment. Despite several advantages offered by AI, some limitations hindered the
widespread applications of these techniques in real water treatment systems. The availability and
selection of data, poor reproducibility and less evidence of applications in real water treatment are some
of the key challenges that need to be addressed. Recommendations are made to ensure the successful
applications of AI in future water-related technologies. This review is beneficial for environmental
researchers, engineers, students, and all stakeholders in the water industry

24
25
Access to clean drinking water is the grand challenge of the modern era and a prime component of
the UN sustainable development goals (SDGs) . On the other hand, water pollution caused by
rapid industrialization and population growth has emerged as a grand environmental challenge in
recent years . Treatment and reuse of wastewater offer a unique opportunity to address both
these challenges. Tremendous progress has been made in the past few decades toward the
development of novel efficient, and cost-effective techniques for the removal of various pollutants
from wastewater . The applications of various optimization and modeling tools have also
gained considerable attention in recent times for assessing performance and improving efficiency.

Artificial intelligence (AI) is the core and well-known branch of computer

the science that deals with building smart systems and resolving problems in a manner comparable
to the human intelligence system. The primary motive of AI applications to a system is to enhance
computer functions that are relevant to human knowledge, such as learning, problem-solving,
reasoning, and perception . AI is a fast-growing field and has real-world applications in diverse
fields such as healthcare, smart cities and transportation, e-commerce, finance, and academia .
AI is further classified into machine learning, deep learning, and data analytics. These techniques are
mainly used for intelligent decision-making, blockchain, cloud computing, the internet of things
(IoT), and the fourth industrial revolution (Industry 4.0) . AI is booming mainly due to its unique
features to learn and adapt a system based on historical data and to

make a decision. AI’s significance is rising incessantly with time due to the integration of AI-based
systems with intelligence, adaptability, and intentionality in their proposed algorithms . AI
systems apply to almost all interdisciplinary fields, and they have played their potential role in
various applications for optimization, classification, regression, and forecasting. AI tools are
sometimes used in combination with experimental design techniques such as response surface
methodology (RSM) to further enhance the precision of optimal solution prediction.

26
The application of AI is emerging in water treatment to overcome the complications of traditional
methods. In the current era, water industries are investing in artificial intelligence, and according to
market research, this investment is expected to reach $6.3 billion by 2030 . Similarly, AI is
expected to save 20 to 30% of operational expenditures by decreasing the cost and optimizing the
usage of chemicals in water treatment . The applications of AI in water treatment have made the
process easy due to its modest implementation, flexibility, generalization, and design simplicity. The
commonly used AI techniques in water treatment

are Recurrent Neural Networks (RNN), Convoluted Neural Networks (CNN), Decision Trees (DT),
Feed Forward Back Propagation Neural Networks (FFBPNN), and Adaptive Network Based Fuzzy
Inference System (ANFIS). The applications of several hybrid techniques such as ANN-GA, MLP-
ANN, ANN-PSO, PSO-GA, Back Propagation (BP)- ANN, Feed Forward Back Propagation
(FFBP)-ANN, and Support Vector Regression (SVR)-GA have also been studied in water treatment.
The availability of data is the main challenge in applications, as AI needs sufficient historical data
to predict future outcomes and offer improvement in the system. Various studies demonstrated the
successful applications of different AI tools for the modeling and optimization of the water treatment
process, such as pollutant removal from water . However, still, various hurdles hinder the
applications of AI in water purification. This review provides a critical analysis of different AI tools
used for assessing the performance of the adsorption process employed for the removal of metals,
dyes, organic compounds, nutrients, pharmaceuticals, drugs, pesticides, and personal care products
(PCPs) from the water. The input variables that affect the process performance are also described,
and the parameters that assess the efficiency of AI models are also discussed. Finally, the significant
challenges in the widespread applications of AI in water treatment are highlighted and
recommendations for future research are also provided.

27
28
AI techniques The most commonly employed AI-based techniques for water treatment are shown in
Fig. 1. These techniques are extensively employed to manage wastewater treatment operations, water
reuse, water-saving, and cost reduction through prediction, diagnosis, assessment, and simulation
.

k-Nearest Neighbour (k-NN) k-NN is a simple machine learning technique used for regression and
classification. k-NN save all the existing data and perform classification on new data points based on
similarity . For example, consider a classification problem having two categories W and Z, as
shown in Fig. 1. Classification of AI techniques. G. Alam et al. Chemical Engineering Journal 427
(2022) 130011 3 Fig. 2. If a new data point occurred, having a placement issue with the W and Z
category, the new data point should be placed in a suitable category based on calculating Euclidean
distance. Therefore, the new point will be added to category Z that have the maximum number of
neighbors. k-NN is the most commonly used technique used for classification problems.

The decision Tree (DT) DT technique is mainly used by AI experts for classification and regression

29
problems. The core purpose of DT is to generate a training model used for class prediction by including
“learning simple decision rules”. It follows a tree structure in which each tree has a node that represents
the attribute or feature of the data, the edge represents the probable answers to a problem, and
the leaf node denotes the real output or class label . This technique is mostly favored because of
its high accuracy and easy implementation. As depicted in Fig. 3, the process may result in many
possible solutions. In a DT technique, all features of a problem are considered from root to leaf node
to detect the optimal solution based on defined conditions Random Forest (RF) RF is used for both
classification and regression problems. Just like the forest, more decision trees mean that robust will
be the RF. It creates DTs on data samples, then makes a prediction on each DT, and lastly, chooses
the optimal solution based on the voting mechanism . The benefit of using RF is that it decreases
the overfitting of the DTs by averaging their result. As shown in Fig. 4; the random samples from a
given dataset are chosen, and a decision tree is built for each sample. Then, the result of each decision
tree is obtained. The next step is to perform the voting process for each predicted result and decide the
most voted predicted result as a final result.

Artificial Neural Networks (ANNs) ANNs are statistical models that are built based on biological
human brain neurons to perform parallel and complex computations. It is used mainly for pattern
recognition problems to

execute modeling and processing nonlinear relationships between the inputsand outputs in a parallel
manner. In ANNs, the neuron represents a node, and the activation functions such as sigmoid and
hyperbolic are used to perform nonlinear computation . ANNs include weights between neurons
(nodes) that can be changed concerning a machine-learning algorithm by using a suitable cost
function to learn from the observed data to improve the model. and consist of many layers, in
which the first layer represents an input layer, the last layer represents the output layer, and thelayers
present between the first and last layers are the hidden layers. An increase in the number of hidden
layers can build complex models that can be trained to improve the performance of ANNs . Fig.
5 shows a simple architecture of ANNs, including the input layer (a, b, c…n), two hidden layers
(hidden layers 1 and 2), and the output layer (a, b…. n). The subtypes of ANNsare discussed below.

30
AI can be leveraged to build efficient water plants and optimize water resources to reduce
energy costs in the long run.

Artificial intelligence is disrupting industries with its wide range of capabilities including augmenting
human intelligence and processing huge data chunks. There have been discussions and reports on
sustainable AI which can work efficiently while conserving the environment. AI has also proved
effective in renewable resources industries. Let us discuss the impact of AI in another sector – the
water sector. Water is an imperative need to live life and it has been going through pollution and
scarcity for a long time. Climate change is a reality that can increase water stress in many places and
increased water contamination will result in a huge water crisis that we are not yet ready to deal
with. According to a report by UNICEF and WHO, 1 in 3 people globally does not have access to
safe drinking water. This scenario is going to become grave in the coming years if we do not address
the issue.

AI in water management might come off as a huge revelation but it can change the way we treat and
manage water sources around us. Let us see how AI can impact the global water sector.

31
8. Artificial Intelligence in Water Treatment and Monitoring

Artificial-intelligence methods and machine-learning models have demonstrated their ability to optimize,
model, and automate critical water- and wastewater-treatment applications, natural-systems monitoring
and management, and water-based agriculture such as hydroponics and aquaponics. In addition to
providing computer-assisted aid to complex issues surrounding water chemistry and physical/biological
processes, artificial intelligence and machine-learning (AI/ML) applications are anticipated to further
optimize water-based applications and decrease capital expenses. Poor data management, low
explainability, poor model reproducibility, and standardization, as well as a lack of academic transparency,
are all important hurdles to overcome to successfully implement these intelligent applications.
Recommendations to aid explainability, data management, reproducibility, and model causality are
offered to overcome these hurdles and continue the successful implementation of these powerful tools.

8.1 Artificial-Intelligence Methods, and Smart Technology

AI models used in water applications are briefly summarized below. A brief mention of utilized AI
methods is also included. A section on smart technologies as defined is included, which are considered
the Internet of Things, smart sensors, and systems based on these technologies, and are often integrated
with AI/ML models and methods. All of these techniques have been studied for uses in water- and
wastewater-treatment processes including chlorination, adsorption and membrane filtration, water-quality
management including dissolved oxygen and water level, as well as water-quality-index modeling and/or
hydroponics and aquaponics farming.

8.1.1. Artificial-Intelligence Methods

AI models and methods are briefly summarized. Their general usages, specific usages in water treatment
and modeling applications, advantages, and disadvantages are highlighted to aid in the selection of
appropriate models and methods for water treatment and monitoring applications. Further published
textbook sources that supply the necessary foundational and in-depth explanations of these methods and
models are also included in the final column. These water treatment and monitoring applications are not
intended to be all-encompassing, but to represent the published journals that were selected based on the

32
methodology explained above. Most of these included ML methods would fit the “black-box” archetype
and would be considered a consistent “disadvantage” for many of the models (notably excluding GA/GPs).

8.1.2. Smart Technology—The Internet of Things (IoT) and Smart Sensing Technology

The Internet of Things is a descriptor for a network of physical objects that can connect to the internet (or
other communication networks) and that are often endowed with some sort of analytical process (such as
environmental sensing) using software, hardware, or other technologies. In the context, the IoT in water
applications often includes internet-enabled systems equipped with pressure sensors, flow sensors, and/or
water-quality/characteristic sensors . The intent is typically to exchange data with other connected
devices or networks over the life and duration of the sensor or other technology , often for system
optimization, transparency, or ease of use . The IoT creates a cooperative network of data collection that
can be stored locally or offsite without a human ever physically needing to take the data themselves or
operate the physical object. As such, the long-lasting function and life of the connected device must be
maintained. Though not technically artificial intelligence, the IoT can be fused with AI to create what has
been coined as the “Artificial Intelligence of Things”, which would marry this data-collection process to
feed AI with critical inputs for its learning process .

Smart sensing technology can be related to IoT, but often represents a broader breadth of systems that do
not need to be defined by their collectiveness and can also include stand-alone or isolated systems/sensors.
To achieve the designation of “smart” sensing technology, the sensors must have some function beyond
their general sensing abilities , which is generally achieved through an actionable decision or
automation. For example, a thermostat that both measures the temperature of a room and interfaces with
a furnace to achieve a set temperature, thus not inherently requiring a connection with other smart devices
in the home. The smart designation can be enhanced through the ability to wirelessly interact with other
systems through Wi-Fi or Bluetooth capabilities.

8.2. Applications in Water and Wastewater Treatment

Artificial intelligence and machine-learning techniques have been studied in several water- and
wastewater-treatment applications. This herein will serve as a cross-section of three common treatment
processes employed at the water- and wastewater-treatment plants. Much of the input data utilized by the
journals were collected and disseminated by treatment-plant staff or other regulatory bodies, relying on
33
traditional collection methods. Using smart-technology integration with AI methods or ML models, the
burden of data collection can be decreased. More data is also likely to increase the accuracy of selected
ML models. Ultimately, this is not intended to represent the gamut of research into AI and ML applications
in the water-treatment industry but rather a representation of current research interest.

AI methods have been demonstrated to be effective in controlling chlorination, while ML models are
effective in modeling DBP concentrations, as well as modeling important parameters for adsorption and
membrane-filtration processes. The results are often evaluated using various statistical measures including
the coefficient of correlation (R), the coefficient of determination (R 2), the mean average error (MAE),
the mean square error (MSE), the root mean square error (RMSE), and relative error (RE).

8.2.1. Chlorination and Disinfection By-Product Management

Disinfection in a water- and the wastewater-treatment plant is the process by which microorganisms and
viruses are killed or inactivated, mainly with chlorine-based disinfectants . While chlorination is
effective as a disinfectant, it also poses human health hazards . Beyond its ability to cause acute toxicity
in humans, chlorine is also known to interact with bromide and organic matter naturally found in water
systems to form what is known as disinfection by-products. Disinfection by-products (DBPs) are
suspected human carcinogens and reproductive disruptors and have received increased scrutiny from
regulators all over the world . DBPs mainly belong to two larger subcategories: trihalomethanes (THMs)
and haloacetic acids (HAAs). THMs are regarded as the most common form of DBPs as their formation
is associated with chlorine disinfectants . Haloacetic acids are commonly tested for five or nine common
haloacetic acids and are commonly referred to as HAA5 or HAA9. The entire mechanism behind the
formation of DBPs in drinking water is not known, making their prediction and mitigation an ideal
candidate for ML technologies. When learning has been achieved, mitigation through control using AI
methods is possible.

Many researchers performed model testing on surface waters that undergo treatment at drinking-water
plants utilizing chlorine as the primary disinfectant, though some studies did involve pre-chlorination
peroxide/ozonation. Researchers also noted success in modeling DBP concentrations in the treated water-
distribution networks, and directly at consumer homes and taps. Common model inputs include water
temperature, pH, chlorine concentration, contact time, and TOC/DOC concentrations. Other successful

34
models have implemented inputs using bromine concentration, UV 254, algae concentrations, chlorophyll-
a concentrations, and DBP-precursor chemical markers.

The most tested ML model used for chlorination and DBP prediction is the ANN, with other applications
involving support vector machines, fuzzy inference systems, and genetic algorithms. In comparative
studies, ANNs typically outperformed both GAs and SVMs, though there are some cases of SVMs
providing a slight advantage when R 2is used as a point of comparison . Common DBPs that were
modeled and/or predicted include total trihalomethanes (TTHM) and total haloacetic acids (THAA), with
some studies focusing on specific DBP compounds including dichloroacetic acid (DCAA), trichloroacetic
acid (TCAA), bromochloroacetic acid (BCAA), HAA5, HAA9, trichloromethane (TCM),
bromodichloromethane (BDCM) and dibromochloromethane (DBCM). Predictions for TTHMs or
THAAs versus their compounds did not differ widely in statistical model-validation numbers.

8.2.2. Adsorption Processes

Adsorption processes are generally regarded as both a physical and chemical treatment option for
removing a wide range of contaminants and pollutants in both the water-treatment and wastewater-
treatment industries. The process of adsorption involves an exothermic mass-transfer surface process that
causes the transfer of some target molecule (or adsorbate) from a fluid to a solid surface (or adsorbent, but
also often referred to as the sorptive media in the industry) . Due to the complexity of the interactions
that affect the efficacy of an adsorptive process , it is often difficult for plants to precisely calculate the
important parameters and ultimate removals of the adsorption process. Reducing this complexity would
enable a treatment plant to extend a sorptive media’s life and increase a treatment plant’s effectiveness
and confidence that it is effectively treating the water according to any applicable rules and regulations.
To further optimize the process, researchers have identified models using ML to make important
predictions for the adsorption process. ML for adsorption processes has the potential to support operator
decisions.

Studies have been published modeling adsorption processes with water streams contaminated with metals,
industrial dyes, and organic compounds. Adsorbent media ranges widely and includes carbonaceous
materials and metal-based nanocomposites, among others. Inputs commonly used during ML modeling of
adsorption processes include pH, water temperature, adsorbent dose, contact time, and initial adsorbate
concentration. Individual models have included inputs utilizing adsorbent particle size, system flow rate,

35
agitation speed, bed height, and BET surface area, among others. Studies that included various organic
pollutants operated with varying compound-specific parameters such as target-contaminant molar mass.
The majority of the published studies that are included pertained to models with outputs of adsorbate
percentage removal (also known as adsorption efficiency), though some models sought to predict
adsorption capacity, non-dimensional effluent concentrations, and the relative importance of input water-
quality parameters.

For studies including metal, organic, and industrial-dye contaminants, the ANN was the most used ML
model. Other models that researchers studied with notable success include ANFIS, SVM, and RF. On
average, ANN, SVM, and RF ML models performed adequately, generally achieving R 2 values greater
than 0.9, and in some cases, greater than 0.99 . In most cases, SVM models performed slightly better
than ANN models, producing both R 2 and RMSE values of better statistical value. In one case, the
optimized ANFIS model performed poorly in comparison to other success models for adsorption
processes, achieving an R = 0.813, and was noted as the worst performing in a comparison between ANN,
ANFIS, and SVM models , though in another it achieved the adequate performance with an R2 = 0.9333 .

8.2.3. Membrane-Filtration Processes

Membrane processes in water and wastewater treatment refer to the separation of contaminants using a
barrier or filter. The water is passed through the membrane usually due to high-pressure differentials
between one side of the membrane and the other side . The smaller the pore size, the more pressure is
required to pass the water through the membrane. Membrane processes are typically used for contaminants
that are difficult or costly to remove by chemical or physical means, but also for contaminants that require
a high level of removal that simply cannot be achieved by other chemical or physical means . The most
used membrane processes are microfiltration, ultrafiltration, nanofiltration, and reverse osmosis.

Researchers have created models that function with microfiltration, ultrafiltration, nanofiltration and
reverse osmosis. A study involving a submerged membrane bioreactor has also been included. Water
sources tested using these models include a wide array of pollutants and natural compounds, including
petroleum/oil, natural organic matter, various industrial and pharmaceutical wastes, and simple salt/ocean
water. Similar to previous sections involving ML in water/wastewater-treatment applications, ANNs are

36
the dominant model used. Other models that have been utilized for membrane-filtration-process modeling
include ANFIS, SVM, and specific forms of ANNs including RNNs (some of which utilize LSTM).

ML techniques for modeling membrane-filtration processes seek to output several different variables,
commonly including transmembrane pressure, permeate flux, and solute rejection. Inputs that exist in
some of these published studies include pH, temperature, contact/filtration time, transmembrane pressure,
and flux rate, among many, and more specific, options. Again, due to the range of models testing for
different parameters, it is difficult to fully compare the statistical values that many of these studies
obtained. Ultimately, ANN, RNN, and SVM models performed adequately well in terms of their
respective R2values, consistently achieving values greater than 0.9, and in many cases, achieving values
greater than 0.99 .

8.3. Applications in Water-Quality Management

Artificial intelligence and machine-learning techniques have been studied in water-quality management.
This here will serve as a cross-section of some water-quality-management models including dissolved
oxygen, among other water-quality parameters and indices, and river-water-level monitoring.

ML models have been demonstrated to be useful for the prediction and modeling of water-quality-
management parameters. The results were commonly evaluated using various statistical measures,
potentially including the coefficient of determination (R 2), the mean square error (MSE), the root mean
square error (RMSE), the normalized root mean square error (NRMSE), the mean absolute percentage
error (MAPE), the Nash–Sutcliffe efficiency coefficient (NSE), the Pearson correlation coefficient (PCC)
and/or accuracy (ACC).

8.3.1. Water-Quality Management

Water-quality management is an important task necessary for the health and good function of aquatic
ecosystems. Often, human activity can hurt the water quality of rivers and other waterways, and tracking
this effect is vital to maintaining these ecosystems. A commonly tracked parameter used to discern the
health of a river or other waterway is the dissolved-oxygen concentration. Hypoxia (or the lack of
dissolved oxygen in waterways) is becoming increasingly prevalent, generally because of increased
nutrient loading and global warming .

37
Due to the interactions between dissolved-oxygen (DO) concentrations and human activity/pollution, it is
increasingly important to measure DO as a means of predicting, and possibly preventing, hypoxic zones
from dealing widespread damage to these aquatic ecosystems. Accurate and real-time results are often
most favorable as moderate decreases in DO represent potentially fatal results in certain species . In
some cases, DO sensing can be obfuscated by environmental factors, demonstrating a present need for
models and methods that can overcome the traditional sensing methods’ shortcomings .

Common inputs for water-quality modeling using ML include pH, water temperature, and BOD levels.
These inputs are also generally the same for water-quality-index (WQI) monitoring and BOD/COD
modeling, with the inclusion of dissolved oxygen as an input in the case of the published studies for WQI
included, while water-level monitoring relies exclusively on past water levels and robust training data.

9. Managing Water Wastage

An India Today report states that it is estimated that around 40% of piped water in India is lost to
leakage. According to a US EPA report, an average family can waste 180 gallons of water per week,
or 9400 gallons of water annually, from household leaks, which is equivalent to the amount of water
needed to wash more than 300 loads of laundry.

We waste a lot of water through leakages, burst pipes, etc. and AI and IoT can help reduce this
wastage. Implementing AI to analyze real-time water loss and automating pipes to shut off whenever
there is a leak can improve the amount of water wastage. AI can predict leaks in storage tanks and
help in mending them before it is too late. Devices connected through IoT can communicate better
and integrate various systems across a city or place

38
Wastewater Treatment

AI can be used to reduce pollutants in the water which in turn decreases water contamination and
scarcity of clean water. AI can be leveraged to detect the amount and composition of toxic
contaminants since AI works on optics, which can increase the efficiency of waste management
systems. Water quality can be continuously monitored and it is possible to get real-time data on the
quality through machine learning and big data. Neural networks and IoT will reduce energy costs
which otherwise increase when using conventional methods.

Digital Water
AI can make the process of water management easier with data analytics, regression models, and
algorithms. These cutting-edge technologies help in building efficient water systems and networks. AI
can be used to build water plants and to get the status of water resources. Water managers and government
bodies can use AI to build a smart water system that can build efficient infrastructure for water
management and can adapt to changing conditions. These systems will be cost-effective and sustainable
that can optimize all water management solutions and predict potential damages

Smart Irrigation

Agriculture is the biggest water-using sector and many lands use a good portion of groundwater for
irrigation purposes. Smart Irrigation will leverage AI systems to minimize the use of water and also
optimize the water resources without wastage. AI systems can detect groundwater levels and also
estimate the agricultural needs to balance the usage of water by guiding sprinkler systems.

More developed precision-based AI systems can predict weather conditions, climate, and humidity
to enable better management of agriculture.

The smart farms will be able to reduce leakages and analyze the soil to determine the condition of
plants and their water needs using AI sensors

39
Artificial Intelligence and integrated management of the water cycle

Water, as a primary source of life and as a natural, cultural, economic, and political resource,
requires intelligent management, and the same artificial intelligence can assist in the involvement
of the collective intelligence dispersed in citizenship, now evolved into a true Smart Communities,
to ensure the protection, conservation and rational and optimal use in an Adaptive Water
Management regime.

The management of water resources requires the formulation of new paradigms capable of
combining, on the one hand, the protection of water resources, through new systems and intelligent
technologies, capable of increasing the efficiency in the use of resources and the performances of
networks and treatment plants present in the territory and on the other hand the development of new
monitoring systems distributed and easy to access for widespread control of the quality status. In
both cases, AI plays an extraordinarily important role, especially in the presence of massive amounts
of data: an increasingly recurring situation due to the strengthening of water and environmental
monitoring systems. The development of interoperable technologies capable of promoting the
dissemination and exchange of large volumes of information between decision-makers, managers,
and citizens, can lead to the creation of widespread knowledge capable of feeding artificial
intelligence
systems and aimed at supporting better environmental protection,ending with a direct impact on the
educational and behavioral side. In this direction, the ubiquity of water, in every declination of social
and productive life, constitutes a natural element to channel information and to consolidate a new
culture that can combine the expressions of AI favoring growth, the sharing of structured expert
knowledge and not and increasing the sense of belonging to one’s territory and to the natural
resources it expresses. The recognition of water, as a human right, passes through the acceptance of
the sense of widespread (public) ownership and responsibility that must guide both the small daily
choices and the big planning, management, political and administrative decisions.

In line with the definitions of the Water Framework Directive 2000/60/EC and the updates in
progress, and in general with the articulated Community, national, and regional regulatory
framework, it is necessary to pursue the objectives of safeguarding, protecting, and improving the
environmental quality of water bodies, as well as the prudent and rational use of natural resources
based on management that is not only sustainable but adaptable to the circumstances that arise also
40
as a result of global changes: all elements of high complexity that find in AI an indispensable ally. In
this vision, participatory processes that can also be activated through AI are crucial for triggering
paths that lead to the construction of the economic and social vocation of smart cities.

A correct understanding of the management of water resources can certainly not be limited to the
simple government of only one of the components such as procurement, distribution networks,
purification, etc., but it requires a broader perspective that allows the analysis and definition of
coordinated and integrated strategies that affect the entire water cycle (Figure 1).

41
Water cycle and pressure factors and areas of application of AI . Legend:
1) climatic change, 2) pollution, 3) physical alterations, and 4) over-exploitation.

Furthermore, even the potentially most efficient strategies have no chance of success if they are not
supported by an “awareness” of citizens who must be directly involved as actors within a system that
cannot ignore virtuous behavior at the macro level and micro-communities.

42
It must be emphasized that the constantly growing demographic evolutions, the consequent
increased use of the intensification of crops, and the effects of climate change with the increase in
the frequency of extreme events, determine an extreme urgency in implementing every possible
solution (including technological and “intelligent”) that can make the resource management system
as a whole more efficient, both in quantitative and qualitative terms.

In this context, the research activities on AI that envisage water management in the implementation
of environmental policies, in close connection with the Europe 2020 strategy which has identified
smart growth, sustainable growth, and inclusive growth, as engines of the relaunch of the economy.

The applications of AI in the water management sector, operating on huge amounts of data, concern
the monitoring and management of extreme events also interface with the “Territory security” area,
while others concern the collection and storage of data, their dissemination, and their interoperable
using interfaces with the “Home automation and Smart Grids” area, in particular concerning aspects
relating to the improvement of the quality of life in domestic environments, the reduction of
management costs and the transmission of information through Power Line Communication (PLC)
and their storage using Cloud technology.

AI can be decisive in identifying and managing adaptation guidelines concerning the climate
changes underway. In particular, it can be useful for:

1. water supply management sector:

2. control of leaks, orienting measurement strategies and priorities, and the most effective types of
intervention to reduce water dispersion;

3. the definition of investments in water networks and infrastructures, supporting a holistic water
policy that takes into account an extremely large number of technical, managerial, social, and
43
 economic variables;

4. water resources management sector:

5. the promotion of the natural conservation of water by orienting the areas in which to favor it both
for employment opportunities and the reduction of hydraulic risk;

6. the aggregation of fragmented surveillance activities between the different management and
control bodies, also to improve the quality and use of information;

7. support capacities in adapting to extreme climatic events, in particular as regards the control of
floods and drought;

As better specified below, in urban and semi-urban areas, AI can also intervene effectively in the
urban wastewater purification sector, orienting technological applications to improve the efficiency
and versatility of plants and favoring low environmental impact technologies, in terms of occupied
surfaces, production of sludge and odor emissions, aimed at maximizing energy recovery and the
recovery of raw materials and in particular nutrients and biofuel

Smart devices in the home: data source for AI applications

The availability of sensors in the home, interfaced via Wi-Fi network to routers or smartphones and
computers, allows to acquire of important amounts of data that can be used both for the benefit of the
individual user, but also and above all for the benefit of the manager, allowing responsible
management of consumption, maintenance of the plants and networks as well as the operating
pressures in the different time bands.

In addition, the intelligent management of water distribution systems allows ample space for the
introduction of innovations in the name of water-saving and environmental sustainability, obtaining
useful advantages in terms of monitoring and optimization of resources.

Technologies based on microelectronic applications make it possible to create multiple systems of

44
 specialized micro and nanosensors, capable of monitoring in real-time the main physic-chemical
parameters that establish the characteristics of the water.

ARTIFICIAL INTELLIGENCE FOR OPTIMIZATION AND DECISION SUPPORT TOOLS IN

WATER SUPPLY

Hydraulic models are used for several purposes:

(i) explanation and prediction tools, for what happened in one location at one time in the water
distribution network without any instrumental and sensor data;

(ii) forecasting tools, for the “what if” scenario for planning and operation of the water distribution
network; and

(iii) prescriptive tools, for decision support platforms, which are becoming popular by advising on
the best options available to solve a particular problem or constraint and, in some cases, automate
the decision.

In its most simple version, decision-making involves optimization techniques to find the combination
of factors that maximize or minimize a numerical objective function.

From the onset of hydraulic modeling, optimization techniques have been applied to water
distribution networks. With the ascent of fast processors and cost-effective sensors resulting in
cheaper computer power, optimization techniques are now mainstreamed into the operation of most
water utilities. One way is via observability analysis, or optimization of the location of a limited
number of sensors to provide the maximum amount of information about the water supply system.
Another is through operational analysis, involving real-time and optimized control of certain parts of
the water distribution network, especially pumping stations and pressure or flow control valves,9 to
reduce energy consumption or monitor water quality.10 In addition to the increased computer and
graphics processing powers, big data generated by social media, mobile phones, and the Internet of
Things (IoT) directly feed into AI, creating a big data ecosystem. The term AI refers to any algorithm

45
that can process data and learn, increasing its performance over time as it becomes better trained.
Many statistical models like decision trees, nonhierarchical classification methods, and Bayesian
networks have become the backbone of machine learning tools, fed by big data.

46
AI in the monitoring of urban waste

The monitoring of discharges assumes particular importance in the context of water management,
both because it is allowed to obtain of useful information along the pipeline to evaluate the presence
of any illegal connections and illegal discharges and because it allows modulating the management
of the plant’s purification as a function of the monitored pollutingload and other ancillary parameters.
The monitoring of the hydraulic efficiency of the drainage system, the chemical and physical
parameters, and the functionality of wastewater treatment plants is an important prerequisite for
ensuring the smooth operation of the environment, health, the city's economy, and society.

47
Artificial intelligence applications for the expeditious characterization of oils in water

In this field too, the main applications of AI derive from the presence of sensors and the progress of
research. Engineers and chemists have made it possible to develop devices capable of evaluating
pollutants present in water such as oils, hydrocarbons, and/or derivatives from a qualitative and
quantitative point of view.

In particular, starting from the acquisition of detailed information on the quality and quantity of
wastewater that passes through a sewer section, it is possible to obtain useful data referring to events
that characterize the functioning of a wastewater collection system (variations in flow rates and load,
anomalous discharges, exceptional meteorological events). Based on the data collected from
chemical/physical monitoring, with the help of AI, it is possible to develop useful knowledge to build
a complete picture of the composition of wastewater, also identifying different parameters and users
(private, artisanal, or industrial) that contribute to the composition of the waste.

48
Problem statement

Let’s consider a basin with J hydropower plants in an urban area, and every hydropower plant has
its water tank. The water depletion, irrigation, and water consumption of all water tanks must be
determined over a specified horizon (R), conditional on operating constraints of hydraulic coupling,
and individual water tanks. The objective is to in- crease the overall hydraulic return. The AI-assisted
decision-making period occurs for 1 or 2 weeks, usually one year for the planning horizon.

It is assumed that discussions will be simplified without losing generality.

(i) The delay in water flow between storages is analyzed using AI-assisted decision-making for long-
term preparation.

(ii) The network of the tank is acyclic as assumed.

(iii) The water is discharged directly from a tank that enters one reservoir alone.

The model actions, states, and dynamics are formulated as follows:

For all r = 0, 1, ...R − 1, the state Y(r) at step r includes the loading yj(t) and the collected water
consumption nj c(t) in period c of water tank j, j =1, ...J. Remember that the leakage has no return
and does not produce energy, and the consumed water is utilized for urban supplies.

Let's consider the control action B(t), and model state Y(t), the system dynamic at step r is identified
as follows: ∀r = 0, 1, … R − 1, j = 1, … J. Furthermore, discussed the dynamic water resource
management issue with annual use in the AIDWRP method.

49
represents the water equilibrium of the cascaded water tanks, where yj(t) is the water tank storage
j at period r and first stage storage yj(0) is assumed as tj(r) The water discharge stated as Eq. 2 isused
to improve the dynamic water resource management issue with annual use in the AIDWRP method.

Eq. 2 indicates the accumulated consumption of every period and lets set the first stage njc(0) = 0
for c = 0, 1, … C − 1. Vj is the set of straight upstream water tank j. The inflows to the water tank j
include the natural inflows, water recession, and water release from theconsumption of an upstream
water tank. ξj(r) Is the natural influx of water tank j at time r, a random parameter with a specified
dispersal. λi The recession ratio of water tank i, which denotes the ratio of consumption returning
to the water model.

3 demonstrates the dynamic planning and control of problem farming.

50
ARTIFICIAL INTELLIGENCE AND HYDRAULIC MODELING 2.0

Hydraulic Modeling 2.0 is the way forward for water utilities to embrace AI at the core of their
planning and operations. Hydraulic Modeling 2.0 represents a qualitative step forward compared to
the conventional Hydraulic Modeling 1.0 currently used by water utilities. The key differences are
summarized in the table. The new paradigm for water distribution network analysis reflected in
Hydraulic Modeling 2.0 draws on the same physical laws and principles in the conventional
Hydraulic Modeling 1.0 models. Therefore, water utilities are expected to move toward the more
advanced Hydraulic Modeling 2.0 based on the experience gained with conventional hydraulic
models. Water utilities considering advanced and high-level hydraulic modeling should first master
the deterministic version of their hydraulic model and identify its limitations as a constraint in their
operations before moving toward Hydraulic Modeling 2.0. According to Global Water Intelligence,
a think tank for the water sector, the potential savings from adopting smart water technologies in the
next decade can amount to around 11% of the total annual expenditures (CAPEX+OPEX) of an
average water utility.15 A 2019 white paper shows that smart water technologies could reduce a water

51
utility’s total cost by 7.4% (baseline costs).16 Most of the savings would come from better and more
efficient CAPEX utilization (12.5% reduction) and improved UFW reduction (3.5%). Although the
AI algorithms’ contribution is not broken down, we can assume that most of the gains come from data
processing and analysis, since sensors and data on their own provide little direct benefit.

52
This study reflects the annual consumption constraint by planning the upper bound consumption
of water in which urban areas help balance the utilization of limited water resources within
various urban regions. Note that the penalty considers this constraint The objective function aims
to increase the predictable overall reward above finite-horizon R

maxπ: ηY0π=G∑t=0T−1ftYrBr+fRYR

As inferred in Eq. where the first stage Y0 is given, policy planning π that contains a sequence
of decision guidelines that

are associated with action space and state space, helps to improve the stochastic dynamics in
Markov Decision Process (MDP), π=π0…πT−1(12)Bt=πtYr, ∀t=0,…R−1

Through the Markov Decision Process, the water resource planning with annual consumption
and locational release has been formulated in this study using. It helps to improve the
locational constraints and develops sensitivity-driven methods for optimizing policies.

53
he proposed the AIDWRP method in an urban water environment.

Sensitivity-based methods are initialized to optimize the Markov Decision Process with action spaces and
discrete states.

As shown where continuous state space Y ∈ ℝm and R is the finite horizon. Let A be the σ field
comprising all the Lebesgue measurable sets. The state Y(r) = y ∈ ℝm at time r, r = 0, 1, … R − 1,
the likelihood that the subsequent state relies on a seta ∈ A at

period r + 1 can be indicated as a transition state expression Qr(A|y)which meets,

Qrℝmy=∫Qrdxy=1,∀y∈ℝm
Let's define linear right operator Qr respective to Qr(A|y) on the function space, To release
locational constraints and develops sensitivity-driven methods to optimize policies in
Markov Decision Process (MDP)

54
FIGURE 9. Artificial Intelligence-based urban water resource management and
planning

10. Challenges and prospects

AI tools have offered serval advantages over conventional mathematical modeling. They can be
used to predict the performance of various water treatment processes and reduce experimental costs.
However, still, some limitations hindered the widespread applications of these techniques in real
water treatment systems. The major drawback of AI tools such as ANNs is the poor reproducibility
due to random weight and bias that might result in a locally optimal solution . The
hybridization of various AI tools can also be employed to predict pollutant removal efficiency
during the adsorption process. Deep learning and deep ANNs are good options for achieving high
accuracy and prediction. However, it requires a sufficient amount of data for experimental training,
testing, finding the local minima, and overfitting. The process performance predicted by AI tools
may also deviate from the actual results under certain circumstances. For example, a sudden change
in operating parameters and water quality may result in wrong predictions by AI tools. Efforts must
be made to strengthen the prediction of AI tools so that they can be employed under various
circumstances and can accommodate sudden fluctuations in the input variables. Based on the
available literature, the AI tools have demonstrated tremendous performance in modeling the batch
adsorption process with a smaller range of data. However, the applications of AI tools in practical
wastewater treatment with a wide range of data are yet to be explored. Another major challenge
relevant to AI-based water treatment is the availability and selection of data. The water utilities are
acquired to generate, collect, process, evaluate and analyze data by creating datasets for system
optimization and prediction. Special attention must be given to selecting the training data for AI
tools, as random data selection is associated with certain drawbacks. However, experimental design
techniques (such as RSM) usually require a large input dataset to create an accurate response. The
operational data from real water treatment plants can be used as input for AI models, and the removal
of pollutants can be predicted more accurately. AI technology could play a critical role in
55
sustainable wastewater treatment and can result in a significant reduction in operating costs in
addition to safeguarding the environment. Besides predicting the efficiency of the water treatment
process, AI tools can be used to integrate the whole process of water treatment starting from water
discharge, transportation, management of sludge, environmental impacts, economy, and
policymaking. Data collection from the various water treatment process is necessary to apply AI
techniques in the water treatment domain successfully. However, special care should be taken while
collecting the data to keep data integrity. All the information, such as data sources, location, process
environment, and dataset ontology, should be listed while reporting the data. This information will
help researchers, students, and engineers to reuse the data in the various experimental domain for
future prediction. AI provides an opportunity for the water industry to optimize and govern water
monitoring and management. The development of new-based algorithms is needed to address certain
problems in water treatment and management, such as water quality, leakage detection, and water
process optimization, to provide intelligent decisions. By applying hybrid AI techniques, prediction
accuracy can be enhanced which leads to a reduction in energy and operational cost. A
benchmark/framework should be developed to compare various AI-based stand-alone and hybrid
techniques in the field of water treatment and to suggest the best techniques for applications in real
water treatment processes.

56
11. Results and discussion

The perspectives of AI in the water sector

AI accelerates the design of systems procurement, distribution, treatment, and reuse of water, using
an increasingly widespread use of computer technology and equipment monitoring. Advanced
diagnostic tools make water management more customized and intelligent in the water sector. In
addition the, in fact, a raised will allow you to overlay information and animations on real-world
images with model projections arising from AI applications to help in activities on daily
management and planning and to manage the resource more efficiently. Virtual reality (VR) can
make “viewable” projections and modeling predictions on the trend basis and patterns of use of
water resources or climatic scenarios assumed, the augmented reality (AR), however, superimposes
information generated by a computer to the real world, in quick time. The AI facilitates the
integration of these worlds, analyses the incoming data stream, manages large information relating
to the scene, and superimposes to do it with big data, images, or animations relevant, also in 3D.
Shortly, we will be able to visualize the system we are imagining to design with the possibility of
visualizing the efficiency of the use of the different scenarios. Engineers, chemists, biologists,
designers, etc. will have new tools to develop collaborative projects, involving expert technicians
and young professionals and evaluating the results of design choices in different scenarios of use.

In the coming years, the offer of AI directed to researchers and companies will be able to expand
further thanks to programs that are simple to use, to be used in the design, promoting the so-called
“fourth industrial revolution”: a systemic transformation that can have direct impacts also in the
management of the waters.
The AI can determine the output of the information correctly exactly at the moment when it is needed,
such as when it is necessary to make choices, reducing the chances of error and, increasing efficiency
and improving productivity.

Contextualizing the water sector, the intervention of the AI will be able to optimize the distribution
or disposal of water, make the removal of polluted substances more efficient , facilitate the reuse
of purified wastewater , provide real-time images of the areas in which criticalities occur.

57
In addition, in the last years, artificial intelligence has started to increase the efficiency of the design
and synthesis phases of new materials that can also be used in the water sector, making applications
faster, easier, and more economical, for example, by reducing the use of chemicals or sludge.

The management of the integrated water cycle is transversal to numerous scientific fields and
artificial intelligence is also expanding in all life sciences because it helps to identify patterns in
complex data sets. The water sector, in particular, allows for huge amounts of data with which
to train algorithms, offering significant development opportunities. In fact, artificial intelligence is
very successful when there is the possibility of a training set of particularly relevant dimensions. The
deep learning and artificial intelligence tools are amazingly powerful that will provide important
answers, especially when interfaced with smart Technologies able to acquire data in multiple areas
of the integrated water cycle.

58
12. Conclusions

AI may soon create a new form of superintelligence life. Nevertheless also in the field of water
management is useful to plan a close relationship synergy between human and artificial intelligence
so that it becomes a useful ally, using a shared ethical approach, transparency of governance of
innovation processes, and the possibility for citizens to exercise their rights and express their
opinions by contributing to the growth of artificial consciousness and collective knowledge.
Markov's decision-making process approach for the management and planning water resources on
a given finite horizon is presented. An agent must define a subset of transfer points and use an MDP
to model the subnetwork's dynamic and uncertainties surrounding those transfer points. The
optimization problem is defined as the reduction of water supply costs subjected to a constraint to
water requirement. The groundwater drawdown can proceed without regulation until the pumping
cost exceeds the user's value of curtailment. The proposed AIDWRP method enhances the
performance and accuracy ratio when compared to other existing methods. The findings indicate
that the artificial intelligence modeling system with the proposed AIDWRP can be utilized to
manage and decision-making in urban water resources management.
Markov's Decision Process in AIDWRP tackles complexes water management with annual usage.
It recognizes local weaknesses that establish sensitivity-driven approaches to optimize multiple
policy areas to allow environmental management and effective ecological planning. Accordingly,
the dedication to water management supply and demand has been alleviated, and significant
increases in local economic performance have been simulated with observed effects.

59