AI in Manufacturing: Market Analysis and Opportunities

Mohamed Abdelaal
[email protected]
Software AG, Darmstadt, Hessen, Germany

ABSTRACT focus on human-machine interfaces (HMI) and AI applications for

In this paper, we explore the transformative impact of Artificial milling machines.
Intelligence (AI) in the manufacturing sector, highlighting its po- The report identifies human-machine interfaces as a key area
tential to revolutionize industry practices and enhance operational where AI is making a significant impact in the manufacturing in-
efficiency. We delve into various applications of AI in manufac- dustry. AI is helping to create more intuitive and user-friendly
arXiv:2407.05426v1 [cs.AI] 21 May 2024

turing, with a particular emphasis on human-machine interfaces interfaces that make it easier for operators to control and monitor
(HMI) and AI-powered milling machines, showcasing how these manufacturing processes. Moreover, the report highlights the use
technologies contribute to more intuitive operations and precision of AI in milling machines. By leveraging AI, milling machines can
in production processes. make real-time adjustments to their operations, leading to more ac-
Through rigorous market analysis, the paper presents insightful curate and precise machining, and ultimately, better products. The
data on AI adoption rates among German manufacturers, compar- report provides insights into how AI is transforming the manufac-
ing these figures with global trends and exploring the specific uses turing industry and how companies can benefit from implementing
of AI in production, maintenance, customer service, and more. In AI in their operations. In this report, we provide a market analysis
addition, the paper examines the emerging field of Generative AI of AI applications in manufacturing. Specifically, we provide a list
and the potential applications of large language models in man- of use cases related to human-machine interaction and the use of
ufacturing processes. The findings indicate a significant increase AI for milling machines. We also answer some questions such as
in AI adoption from 6% in 2020 to 13.3% in 2023 among German how many companies in Germany use AI? How many of those
companies, with a projection of substantial economic impact by companies are in manufacturing? Which AI application is the most
2030. The study also addresses the challenges faced by companies, common in manufacturing?
such as data quality and integration hurdles, providing a balanced In summary, this article makes significant contributions to the
view of the opportunities and obstacles in AI implementation. understanding of AI adoption and its potential in the manufacturing
industry. By providing a comprehensive market analysis, the report
ACM Reference Format: offers valuable insights into the current state of AI implementation,
Mohamed Abdelaal. 2018. AI in Manufacturing: Market Analysis and Op- the challenges faced by manufacturers, and the successful use cases
portunities. In Proceedings of Make sure to enter the correct conference title
that demonstrate the benefits of AI-driven solutions. We break down
from your rights confirmation emai (Conference acronym ’XX). ACM, New
the manufacturing process into its constituent phases, identifying
York, NY, USA, 10 pages. https://doi.org/XXXXXXX.XXXXXXX
key areas where AI can be integrated to optimize performance
and efficiency. Additionally, we explore the role of human-machine
1 INTRODUCTION interfaces and the potential of Generative AI and large language
Manufacturing has been one of the early adopters of artificial intel- models in revolutionizing various aspects of manufacturing. By
ligence (AI) and has witnessed significant benefits from the tech- presenting a detailed examination of AI’s transformative potential,
nology’s implementation. With the use of AI, manufacturers can this article serves as an essential resource for stakeholders in the
analyze vast amounts of data generated by sensors, machines, and manufacturing sector, guiding them toward a more competitive
other sources to identify inefficiencies, predict maintenance needs, and innovative future.
and optimize production schedules. Additionally, AI can broadly im- The remainder of the article is structured as follows. Section 2
prove safety in manufacturing environments by analyzing data in delves into the adoption of AI by German manufacturers, present-
real-time and identifying potential safety hazards before they cause ing survey data from various sources and discussing the challenges
accidents. By adopting AI technologies, manufacturing companies faced by companies in implementing AI technologies. Section 3
can gain a competitive advantage in the market and stay ahead of provides an overview of the typical phases of the manufacturing
the curve in an ever-changing business landscape. In this report, process, setting the stage for the subsequent discussion of AI use
we explore the potential of AI in manufacturing, with a particular cases. Section 4 forms the core of the article, presenting a compre-
hensive list of AI use cases in manufacturing, supported by real-
Permission to make digital or hard copies of all or part of this work for personal or world examples of German companies successfully implementing
classroom use is granted without fee provided that copies are not made or distributed these solutions. Section 5 is dedicated to the role of AI in enhanc-
for profit or commercial advantage and that copies bear this notice and the full citation ing human-machine interfaces, before Section 6 introduces the
on the first page. Copyrights for components of this work owned by others than the
author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or potential of AI for milling machines. The potential of Generative
republish, to post on servers or to redistribute to lists, requires prior specific permission AI and large language models in manufacturing is explained in
and/or a fee. Request permissions from [email protected].
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY
Section 7. Finally, Section 8 concludes the article with a summary
© 2018 Copyright held by the owner/author(s). Publication rights licensed to ACM. of its contributions.
ACM ISBN 978-1-4503-XXXX-X/18/06
https://doi.org/XXXXXXX.XXXXXXX
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Abdelaal et al.

Marketing 71% production domain are expected to generate an additional 430 bil-
Production & Maintenance 64%
Customer Service 63% lion euros from AI applications. As a result, the gross domestic
Sales 53% product at the end of the decade will be 11.3 percent higher than
Reporting or Translation 50% it would have been without using this technology, according to a
Accounting 44%
Strategy Development 43% study by the consulting firm PricewaterhouseCoopers [25]. The
IT Department 39% survey by Capgemini demonstrated that AI applications used at
Logistics 35% German companies in the manufacturing sector include process
Research & Development 25%
Human Resource 21% and logistics data analysis, quality assurance, machine control, data-
Controlling 18% driven process modeling, online condition monitoring, anomaly
Legal & Tax 1%
0 20 40 60 80 100 detection, adaptive control systems, and simulation technology.
Percentage Aside from the BitKom’s survey, there exist also other surveys
with distinct conclusions. For instance, the survey by Statista in
Figure 1: AI adoption in German companies (adapted 2019 found that 23% of German industrial companies with 500 to
from [31]) 1,999 employees used AI in the context of Industry 4.0 practices [17].
A more recent survey conducted by Harris Poll on behalf of Google
Cloud in 2021 found that 79% of manufacturers in Germany are
leveraging AI on a daily basis, with quality control and supply
2 AI-ADOPTION BY GERMAN chain optimization being the two main application areas for AI
MANUFACTURERS in manufacturing [37]. Such a survey revealed that the top three
In this section, we explore the various studies of AI involvement manufacturing sub-sectors deploying AI to assist in day-to-day
in German firms, putting more focus on the manufacturing sector. operations are automotive/OEMs with 76%, automotive suppliers
There exist several surveys and reports that investigated the adop- with 68%, and heavy machinery with 67%. Finally, a recent survey
tion of AI in German manufacturing. For instance, Statista recently conducted by ZEW in 2023 revealed that only 10% of manufacturing
released a report about AI adoption in Germany [1]. The report companies and 30% of companies in the information economy in
states that the AI market is projected to reach a size of $7.85 billion Germany had implemented AI [38]. Additionally, another quarter
in 2024. It is expected to grow at an annual rate (CAGR) of 28.41% of companies in both sectors planned to adopt AI in the future.
from 2024 to 2030, resulting in a market volume of $35.19 billion However, only a small percentage of companies reported having a
by 2030. According to the ifo Business survey in 2023, 13.3% of high level of competitiveness in AI: only 6% in the manufacturing
German companies are currently using AI, and 9.2% are planning industry and 15% in the information economy.
to adopt it. Additionally, 36.7% of all companies surveyed are ex- Despite its potential, German manufacturers often encounter nu-
ploring potential use cases for AI [14]. Such results show a better merous challenges in the successful and effective implementation
position of AI adoption than the survey conducted by the digital of AI in their production environments. One of the main challenges
association Bitkom in 2021 which states that AI applications are is the need for clean, meaningful, high-quality data, which is critical
used in only 8% of German companies, compared to 6% in 2020 for the success of AI initiatives but can be a challenge in manu-
[31]. Nevertheless, nearly 70% of German companies believe AI is a facturing. Biases in manufacturing data can be caused by multiple
key technology of the future. The Bitkom’s survey also examined factors, such as unrepresentative or incomplete training data or the
what AI is used for in the German companies that have already reliance on flawed information. For example, collecting data from
implemented this technology. German companies predominantly a specific group of workers may lead to biased data that does not
use AI for marketing purposes, with personalized advertising be- reflect the entire workforce. Additionally, biases can occur when
ing the most common application (71%). Furthermore, 64% of these data is collected from a specific time period, leading to outdated
businesses utilize AI to optimize their internal production and main- data that does not reflect current conditions. Other challenges stem
tenance processes. Approximately 63% of companies integrate AI from the lack of business alignment, difficulty assessing vendors,
into their customer service, utilizing it to automatically answer integration challenges, and legal issues. Furthermore, Germany lags
queries. Additionally, AI is employed by about 53% of companies behind other countries in developing AI applications, which may
to analyze customer behavior in sales and by 50% of companies to make it more difficult for German manufacturers to implement AI
prepare text-based materials such as reports or translations across in their operations [16]. Figure 2 demonstrates a set of barriers to
various departments. Figure 1 delineates the distributions of AI AI in German small and medium enterprises (SMEs), which have
applications adopted by German companies. The figure highlights been identified in an empirical study in 2021 [34].
that the highest percentage of companies using AI is in marketing,
followed by production and maintenance, customer service, and
sales. 3 MANUFACTURING PROCESS
Focusing on production, we found that the results obtained in In this section, we elaborate on the AI use cases in manufactur-
Bitkom’s survey broadly conform with the results obtained in an- ing. Before delving into the use cases and for the report to be
other survey by Capgemini in 2019. Specifically, the survey carried self-contained, we provide a glimpse into the typical phases of the
out by Capgemini showed that 69% of the manufacturers in Ger- manufacturing process. Subsequently, we explore the possibility of
many use AI in their day-to-day operations, compared to 28% in integrating AI into the production process, providing a list of spe-
the US and 11% in China [5]. By 2030, German companies in the cific use cases for AI in each phase. In general, the manufacturing
AI in Manufacturing: Market Analysis and Opportunities Conference acronym ’XX, June 03–05, 2018, Woodstock, NY

Lack of Competence 65% In the detailed planning phase, the project plan is finalized. This
Obstacles at Implementation 52% involves final time management, commissioning, and other activi-
Data Problems 52%
Deficiencies in IT 46% ties necessary to get the project ready for production. The company
Financial Barriers 39% must ensure that all steps in the production process are identified
No Commitment 32% and scheduled, including quality control checks and any necessary
Regulatory Hurdles 32% equipment maintenance. The commissioning process is essential
No Business Case 28% in ensuring that the final product meets all requirements before
Cyberattacks 13%
No Value Added 7% delivery. During this phase, the company must ensure that all stake-
Other 7% holders are aware of the project plan and timelines to ensure ev-
0 20 40 60 80 100 eryone is on the same page. The production phase is where the
Percentage actual manufacturing and assembly of the product take place. It
includes transporting the necessary materials and components to
Figure 2: Barriers to AI in German SMEs (adapted from [34]) the production site and assembling them according to the project
plan. Post-work or improvement is also carried out during this
phase to ensure that the product meets all quality requirements.
The company must ensure that all production activities are care-
Work Detailed
Order Intake fully monitored to avoid any quality issues that can arise during
Preparation Planning
Customers the manufacturing process.
The quality assurance phase is where the company carries out a
series of quality checks on the product to ensure it meets all quality
Maintenance & Quality
Delivery Production requirements. This includes checking for defects, ensuring that
Servicing Assurance
the product meets all specifications, and conducting any necessary
tests. The company must ensure that all quality checks are carried
Figure 3: Typical phases of the manufacturing process out according to the quality control plan to avoid any issues that
can affect the final product’s quality. The delivery phase involves
shipping the product to the customer or storing it for future use.
process can be interpreted as a chronological sequence of several This involves coordinating with the shipping company to ensure
activities, including sourcing planning, resource management, per- that the product is delivered according to the customer’s specifi-
sonnel planning, and optimization of processes in inbound and cations. The company must also ensure that the product is stored
outbound logistics. Figure 3 depicts a set of typical phases of the appropriately if it is not immediately required by the customer. This
manufacturing process. phase is critical to ensuring that the customer receives the product
The manufacturing process typically begins with the order intake on time and in excellent condition. The maintenance phase is the
when the company receives the customer’s request for a specific final stage of the production process. This involves maintaining the
product or service. The order intake process involves several steps, product, ensuring that it is functioning correctly, and addressing
including capturing the customer’s requirements, determining the any issues that arise. The company must ensure that all mainte-
feasibility of the project, and providing the customer with a quote nance activities are carried out according to the maintenance plan,
or estimate. During this phase, the company must establish clear including any necessary repairs or upgrades. The maintenance
communication with the customer to ensure they fully understand phase is critical to ensuring that the product continues to meet the
their requirements and are satisfied with the proposed solution. customer’s requirements and operates efficiently.
The accuracy and completeness of the order intake process can
impact the rest of the production process and ultimately affect
customer satisfaction. Therefore, it is essential to carefully manage 4 AI USE CASES IN MANUFACTURING
this phase to ensure that all customer requirements are met and In this section, we discuss the potential of AI in the manufacturing
properly communicated to the production team. process, with providing examples of real-world use cases of the AI
In the work preparation phase, the production team analyzes adoption by German manufacturers.
the order details and plans how to complete the project. Resource
management, supply chains, and scheduling are the key areas to
focus on during this phase. The company must ensure that the nec- Chatbots. In fact, AI-powered chatbots can be used to interact
essary resources, including personnel, materials, and equipment, with customers and gather their order details, which can reduce the
are available to complete the project. The company must also con- workload on human employees and free them up to focus on other
sider the supply chain, ensuring that all necessary components and important tasks. The chatbots are not only used as a communication
materials are available when needed to avoid production delays. tool with the customers but they can also be leveraged as an effective
The scheduling of the project is critical to ensure timely delivery, tool for facilitating repeated tasks done by the employees. As an
and it involves coordinating various production activities to ensure example, the Avanade chatbot adopted by Siemens [4] has improved
that everything happens according to plan. The work preparation the overall digital workplace experience for Siemens employees. It
phase lays the foundation for the success of the entire production saves them time searching for up-to-date information and getting
process. answers to repetitive questions.
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Abdelaal et al.

Digital Twins. In general, digital twins are virtual representa- algorithms and image recognition to power its fleet of self-driving
tions of physical objects or systems that span their lifecycle, are ships, which in turn improves its supply chain efficiency and safely
updated from real-time data, and use simulation. The integration transports its cargo [23].
of AI and machine learning algorithms with digital twins allows
businesses to quickly analyze and interpret data to make informed Predictive Maintenance. ML algorithms for predictive mainte-
decisions [21]. This integration of AI improves digital twin technol- nance can help manufacturers identify potential equipment failures
ogy by enabling businesses to obtain granular insights into their before they occur. To this end, AI-driven technologies have the ca-
operations, enabling them to optimize performance, reduce down- pability to gather and analyze vast amounts of data, such as audio,
time, and improve product quality. AI algorithms can be used to video, and GPS, obtained from sensors situated on the factory floor.
analyze large amounts of data generated by digital twins to identify These technologies can identify abnormalities or inefficiencies in
patterns, predict outcomes, and optimize performance. In 2020, the equipment, which can forestall unexpected machinery breakdowns.
worldwide digital twins market was worth $3.1 billion. However, it Such anomalies could take the form of unusual sounds emanat-
is projected to surpass $48.2 billion by 2026 [32]. ing from a vehicle’s engine or a malfunction in an assembly line.
German manufacturers have been at the forefront of digital twin Accordingly, AI technologies can enhance factory efficiency and
technology and AI integration. Siemens, for example, has developed reduce costs by avoiding unscheduled equipment downtime. Ac-
a digital twin of a gas turbine that uses AI algorithms to optimize cording to a report by McKinsey, the most significant benefit of AI
its performance [15]. Bosch has developed a digital twin of a pro- in the manufacturing industry is through predictive maintenance,
duction line that uses AI to predict maintenance needs and reduce which has the potential to generate global value between 0.5𝑡𝑜0.7
downtime. Volkswagen VW and Bosch are using data from the Golf trillion [23]. BCG has labeled predictive maintenance as the pri-
8 in Europe and a digital twin model to improve the performance mary Industry 4.0 focus, with particular importance for cement
of self-driving cars by simulating the performance of the car in dif- manufacturers.
ferent driving conditions [11]. The digital twin of the BMW factory
replicates every aspect of the production process, including avatars Quality Control. A system capable of detecting defects that may
of workers transporting equipment along the production lines, tech- go unnoticed by the human eye and triggering corrective measures
nicians shaping the metal frame of each vehicle, and quality control automatically can broadly help reduce product recalls and minimize
experts thoroughly inspecting each component [32]. wastage. Additionally, the system can detect anomalies such as toxic
gas emissions in real time, which can prevent workplace hazards
Data Analysis. AI algorithms can analyze customer data and
and enhance worker safety in factories. In this context, computer
purchase history to make personalized recommendations, which
vision for quality control can help manufacturers detect defects
can increase the chances of customers making additional purchases.
and anomalies in products and processes. For example, the BMW
Additionally, AI can help automate the order-taking process by
Group uses automated image recognition for quality checks, and
automatically processing orders and generating invoices, reducing
inspections, and to eliminate pseudo-defects (deviations from target
the time and effort required to complete these tasks manually [2].
despite no actual faults) [23]. As a result, they have achieved high
For instance, Thyssenkrupp Materials Services has developed an AI-
levels of precision in manufacturing. Another company that’s ben-
powered platform called “alfred” that uses ML algorithms to analyze
efited from AI in manufacturing is Porsche. They use autonomous
customer data and make personalized recommendations for each
guided vehicles (AGVs) to automate significant portions of auto-
customer [33]. The platform has helped to increase customer loyalty
motive manufacturing. The AGVs take vehicle body parts from
and improve the efficiency of the order-taking process.
one processing station to the next, eliminating the need for human
Supply Chain Optimization. AI can help manufacturers optimize intervention and making the facility resilient to disruptions like
their supply chains by analyzing data on suppliers, inventory levels, pandemics.
and customer demand. For example, AI algorithms can analyze his-
torical data on supplier performance and delivery times to identify Factory Automation. By leveraging AI-driven process mining
potential bottlenecks and proactively manage inventory levels. This tools, manufacturing companies can automatically identify and
can help to reduce lead times and improve customer satisfaction. remove production process bottlenecks, while also facilitating per-
Moreover, AI-enabled systems for supply chain management can formance comparison across different regions. This, in turn, enables
help manufacturers assess various scenarios to improve last-mile them to standardize and streamline workflows to create better man-
deliveries, predict optimal delivery routes, track driver performance ufacturing processes. Another use case involves robotic process
in real-time, and assess weather and traffic reports besides historical automation, where robots independently execute repetitive tasks
data to forecast future delivery times accurately. McKinsey predicts on the factory floor, with human intervention being necessary only
that AI-enhanced supply chains will reduce forecasting errors by in case of exceptions or anomalies. Additionally, robots can em-
20-50%, reduce lost sales by 65%, and reduce the over-stocking in- ploy computer vision to screen and examine processes without any
ventories by 20-50% [23]. As an example, BMW has implemented an human intervention. Accordingly, robotics and automation for fac-
AI-powered supply chain optimization system that uses ML algo- tory operations can help manufacturers increase efficiency, reduce
rithms to analyze data on suppliers, inventory levels, and customer costs, and improve safety. For instance, McKinsey suggests that the
demand [10]. The system has helped BMW to broadly reduce inven- implementation of AI in the semiconductor industry for process
tory costs and improve on-time delivery rates. Another example automation can increase yields by up to 30%, while also reducing
is that of the carmaker Rolls Royce, which leverages advanced ML testing costs and scrap rates [23].
AI in Manufacturing: Market Analysis and Opportunities Conference acronym ’XX, June 03–05, 2018, Woodstock, NY

AI-Powered Design. AI-enabled software can help create several processes, and they are usually found on machines that require
optimized designs for a single product. The software, also known constant feedback from the system or printed production reports.
as generative design software, requires engineers to provide certain Another way to categorize HMIs is based on their level of control.
input parameters such as raw materials, size and weight, manu- There are two basic types of HMI: supervisory level and machine
facturing methods, and cost and other resource constraints. Using level [6]. On the one hand, supervisory-level HMIs are usually
these parameters, the algorithm can generate various design per- designed for control room environments and used for system-wide
mutations. The software lets engineers test various designs against monitoring and control. On the other hand, machine-level HMIs
a wide collection of manufacturing scenarios and conditions to use embedded, machine-level devices within the production facility
pick the best possible outcome. The carmaker Nissan is using AI itself. In addition to these types, there are also touch-screen HMIs,
to develop never-seen-before car designs in the blink of an eye. mobile HMIs, and web-based HMIs. Touch screen HMIs are used
The process would take human designers months, or even years to to control and monitor machines and processes through a touch
complete. Such software can also be used to pick the right recipes screen interface. Mobile HMIs are used to control and monitor
that lead to the least amount of raw material and energy waste [23]. machines and processes through a mobile device. Web-based HMIs
are used to control and monitor machines and processes through a
Resource Management. AI can help manufacturing firms manage
web browser.
their resources more efficiently by analyzing historical data on re-
source utilization and predicting future demand. For example, AI
algorithms can analyze data on machine usage and downtime to
optimize scheduling and reduce idle time. This can help to improve 5.2 AI-Adoption to HMI Systems
production efficiency and reduce costs. Since they collect data in
In general, HMIs can allow humans to interact with machines using
real-time, manufacturers can monitor their warehouses continu-
natural language, gestures, and other intuitive methods, and AI can
ously and plan their logistics better. Demand forecasting can further
interpret and respond to these inputs. There exist currently several
help manufacturers take action to stock up their warehouses in
examples of companies that are using HMIs to create new business
advance and keep up with the customer demand without enormous
models, such as healthcare companies using virtual assistants to
transportation costs. Robots in the warehouses can track, lift, move,
provide medical advice and financial companies using chatbots
and sort items, leaving the more strategic tasks to the humans
to assist customers with banking transactions [13]. In this way,
and reducing workplace injuries. Automated quality control and
HMIs can improve efficiency, reduce costs, and increase customer
inventorying can reduce warehouse management costs, improve
satisfaction. Therefore, many companies nowadays explore ways
productivity, and require a smaller labor force. As a result, manu-
to incorporate this technology into their operations.
facturers can increase their sales and profit margins [15]. AI can
Traditionally, HMI solutions used to be independent, separate
help manufacturing firms optimize their production schedules by
terminals that were installed by an OEM (Original Equipment Man-
analyzing data on machine usage, labor availability, and customer
ufacturer) as a component of a machine. However, the latest HMI
demand. For example, AI algorithms can analyze historical data
solutions that utilize Machine Learning have been designed to au-
on production times and machine availability to create more accu-
tomatically transmit data to either an on-premise solution or the
rate production schedules. This can help to reduce lead times and
cloud, without any need for configuration. Nowadays, companies
improve on-time delivery rates.
expect HMI solutions equipped with some Machine learning algo-
rithms that can learn from all the data generated from IoT sensors
5 HUMAN-MACHINE INTERFACES and adapt to the ongoing behavior of the operator. However, there
In this section, we introduce the human-machine interfaces and exist several challenges to implementing AI-powered HMIs, such
how they work. Subsequently, we discuss several ideas for the ap- as privacy concerns and the need for high-quality data to train AI
plication of AI technologies to optimize the performance of human- models.
machine interfaces. According to recent industry reports, the HMI (Human Machine
Interface) market was valued at USD 4352.7 million in 2021 [9]. The
5.1 What are Human Machine Interfaces? HMI market is predicted to witness significant growth, with an
In general, human-machine interfaces (HMIs) serves as a mediator expected market size of USD 9652.6 million by the year 2030. This
between a machine and the operating personnel, where it connects growth can be attributed to the increasing demand for advanced
an operator to the controller for an industrial system. They typically HMI solutions in a variety of industries, such as manufacturing,
include visual representations of the systems. Through an HMI, an automotive, and healthcare, among others. The adoption of HMI
operator can control the operation of machinery and associated solutions has been driven by the need for more intuitive and user-
devices in industrial environments. There are different types of friendly interfaces that enhance productivity and efficiency while
Human Machine Interface (HMI) available in the market. The most reducing errors and downtime. Additionally, the rapid advance-
common types of HMI are push-button, overseer, and data-handling ments in technology, including AI and IoT, are expected to further
HMIs [8]. Push-button HMIs are used to control simple machines boost the growth of the HMI market in the coming years.
and processes, and they are usually found on small machines. Over- There are several companies that are currently using AI to cre-
seer HMIs are used to monitor and control complex machines and ate HMI solutions. The key players in the market include Siemens
processes, and they are usually found on large machines. Data han- AG, Schneider Electric SE, ABB Ltd., and Rockwell Automation,
dling HMIs are used to collect and analyze data from machines and Inc., Samsara. For example, Samsara offers an HMI that empowers
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Abdelaal et al.

operators with real-time asset health data and streamlines opera- AI can enhance the security of devices by enabling them to detect
tions to enable a proactive approach to maintenance [28]. Another and respond to threats in real-time, without relying on cloud-based
company is Schneider Electric, which provides a comprehensive solutions.
HMI solution that manages all vital machine features, including
visualization, control, supervision, diagnostics, monitoring, and
data logging. Schneider Electric’s HMI solution is designed to be 6 AI FOR MILLING MACHINES
highly connectable and flexible, allowing for easy collaboration be- In this section, we explain the milling machines and their com-
tween operators and machines. Additionally, Rockwell Automation ponents and applications. Afterward, we discuss the potential of
is another company that is using AI for creating HMI solutions. leveraging AI technologies to develop efficient and safe milling
Rockwell Automation’s HMI solution uses AI to provide predictive machines.
maintenance and real-time analytics to improve machine perfor- In general, a milling machine is a type of machinery used to
mance and reduce downtime. remove material from a workpiece using rotary cutters [19]. The
Renesas offers a scalable SMARC 2.1 System on Module (SoM) process of removing pieces of the material in line with the tool
with an AI solution based on the Renesas RZ/G2 microprocessor axis is known as milling, hence the name. Milling machines can
family and a full power and timing tree [26]. This HMI SoM with drill, bore, and cut an array of materials and come in many types,
AI accelerator is designed for AI-enabled human-machine inter- used across various industries. There are several types of milling
face (HMI) applications, such as smart industrial control terminals. machines available, including computer-controlled vertical mills
The HMI SoM with AI accelerator is a turnkey architecture that with the ability to move the spindle vertically along the Z-axis.
provides a comprehensive solution for HMI applications, including Milling can be done with a wide range of machine tools, and it is
visualization, control, supervision, diagnostics, monitoring, and one of the most commonly used processes for machining custom
data logging. Similarly, Microchip offers a range of smart human- parts to precise tolerances. The original class of machine tools
machine interface (HMI) solutions that use machine learning to for milling was the milling machine (often called a mill). Disk- or
enhance the user experience and improve efficiency [24]. Their barrel-shaped cutters are clamped through holes in their centers to
solutions include touch and touchless sensing, gesture recognition, arbors attached to the machine spindle; they have teeth on their
and voice control. These technologies can be used in a variety of peripheries only or on both peripheries and faces. An end mill is a
applications, including automotive, industrial, and medical devices. cutter shaped like a pencil with a tapered shank that fits into the
By leveraging machine learning, Microchip’s HMI solutions can machine spindle; it has cutting teeth on its face and spiral blades
adapt to the user’s behavior and preferences, enabling more intu- on the lateral surface.
itive and natural interactions with machines. Overall, Microchip’s AI has several applications in milling machines. Specifically, it
HMI solutions offer a reliable and innovative way to enhance the has the potential to enhance production capacity, boost machine
usability and efficiency of human-machine interfaces. productivity, and optimize the CNC (computer numerical control)
Technically Speaking, AI and one-shot learning can revolution- machining industry by improving accuracy, reducing redundant
ize the field of human-machine interaction by enabling more intu- tasks, predicting maintenance needs, optimizing parameters, and
itive and natural interfaces that adapt to the user’s behavior and improving surface quality [30]. Here are some ways in which AI
preferences [12]. One-shot learning can enable machines to learn and machine learning are being applied to CNC machining [35]:
from a single example, reducing the need for extensive training
data and making it easier for machines to adapt to new tasks and • Cutting forces and cutting tool wear prediction: Machine
environments. The combination of AI and one-shot learning can learning systems can be applied to predict cutting forces and
enable machines to understand human emotions and intentions, cutting tool wear in CNC machine tools to increase cutting
enhancing the user experience and enabling more efficient and tool life during machining operations.
personalized interactions. The use of natural language processing • Optimized machining parameters: Advanced machine learn-
(NLP) and voice recognition can enable more natural and intuitive ing systems can be used to obtain optimized machining pa-
interactions between humans and machines, reducing the need for rameters of CNC machining operations in order to increase
physical interfaces and touchscreens. The integration of AI and one- efficiency during part manufacturing processes. For instance,
shot learning in human-machine interaction can have significant such parameters can be tested within a digital twin —a vir-
implications for a wide range of industries, including healthcare, tual copy of the CNC machine—using machine learning and
automotive, and manufacturing. IoT to optimize CNC machinery.
To unleash the potential of AI for HMI solutions, edge AI tech- • Surface quality prediction: Surface quality of machined com-
niques are to be adopted [29]. In general, Edge AI brings intelligence ponents can be predicted and improved using advanced ma-
closer to the devices, reducing the latency and improving the re- chine learning systems to improve the quality of machined
sponse time of HMIs. With edge AI, HMIs can adapt to the user’s parts.
behavior and preferences, enabling more personalized and intu- • Machine scheduling: AI can be used for machine scheduling
itive interactions. Edge AI-powered HMIs can enable devices to for optimized performance and downtime, and better data
understand and respond to voice commands, reducing the need analysis, programming, and testing.
for physical interfaces. Edge AI can enable HMIs to analyze data • Milling diagnosis: AI approaches can be used for milling diag-
from sensors and other sources in real time, enabling predictive nosis in composite sandwich structures based on honeycomb
maintenance and improving the overall reliability of devices. Edge core.
AI in Manufacturing: Market Analysis and Opportunities Conference acronym ’XX, June 03–05, 2018, Woodstock, NY

• Reduction of redundant tasks: AI can reduce redundant tasks • CAD/CAM data: Computer-Aided Design (CAD) and Computer-
for manufacturers, which significantly impacts the time re- Aided Manufacturing (CAM) data can be used to train models
quired for machining and human effort and makes the parts for tasks related to machining toolpath optimization, tool
more identical to one another. selection, or simulation. CAD/CAM data provides geometric
The utilization of modern generative models, e.g., OpenAI GPT4 information about the workpiece, tooling, and machining
and Google BARD, has the potential to revolutionize CNC program- operations, enabling the models to learn from simulated or
ming by generating highly optimized and efficient programs [3]. historical machining processes.
With their advanced natural language processing capabilities, the • Historical production data: Historical data from past CNC
generative models are capable of comprehending intricate machin- machining operations, such as machining parameters, work-
ing specifications and producing G and M codes tailored precisely piece characteristics, and quality metrics, can be used to
to meet project requirements. This groundbreaking AI solution train models for tasks like predictive maintenance, process
significantly saves time and effort during the programming pro- optimization, or estimating machining time.
cess while minimizing the potential for errors inherent in manual • Expert knowledge or rules: In addition to data-driven ap-
programming methods. By incorporating generative models into proaches, expert knowledge or domain-specific rules can
CNC programming workflows, manufacturers can benefit from the be incorporated into machine learning models. These rules
following advantages: can be derived from expert machinists’ experience or best
practices in CNC machining.
• Accelerated program generation: the generative models can
rapidly generate CNC programs, enabling manufacturers to
The market for milling machines is projected to grow signifi-
expedite production timelines and meet demanding dead-
cantly in the coming years, with the adoption of predictive analytics
lines.
driving growth [36]. The global market for milling machines is pro-
• Improved program quality: Leveraging AI-driven algorithms,
jected to reach $10.6 billion by 2030, growing at a CAGR of 5.6% over
the language model analyzes historical data to generate op-
the analysis period 2022-2030. The market growth will be aided by
timized programs that enhance machining efficiency and
the rising demand for customized mass production, as the milling
diminish tool wear.
machine produces customized products in large quantities. The
• Enhanced adaptability: the generative models’ advanced al-
micro-milling machine market has seen significant growth in re-
gorithms enable it to adapt to evolving machining demands,
cent years, driven by increased demand for its products and services.
ensuring that generated programs remain up-to-date and
The integration of additive manufacturing and artificial intelligence
aligned with the latest requirements.
is propelling the market growth of micro-milling machines.
• Reduced reliance on skilled programmers: By minimizing the
There are several companies that develop artificial intelligence
need for extensive human intervention, the generative mod-
CNC milling machines, including DMG Mori, TRUMPF, Mazak, ES-
els broadly reduce manufacturers’ reliance on highly skilled
PRIT, and Okuma [3]. For instance, DMG Mori, a renowned machine
CNC programmers, allowing for more efficient allocation of
tool manufacturer, is at the forefront of integrating AI systems into
resources.
their CNC machines. The company is dedicated to enhancing ma-
In general, machine learning models for CNC machining can chine performance, boosting productivity, and minimizing errors
be trained using various types of data, depending on the specific in the manufacturing process through their AI-powered solutions.
task or objective. Common types of data used for training machine DMG Mori’s AI-driven technologies encompass intelligent moni-
learning models in CNC machining include: toring systems that optimize the machining process. Furthermore,
• Sensor data: CNC machines often have sensors that collect their advanced AI algorithms automatically adapt cutting parame-
data during machining operations. This data can include ters to minimize tool wear and prolong the lifespan of tools.
information such as cutting forces, spindle speeds, tool tem- Mazak, another prominent player in the CNC industry, is ac-
peratures, vibrations, and power consumption. Sensor data tively driving the transformation of manufacturing through the
provides valuable insights into the machining process and development of AI-powered technologies. One notable innovation
can be used to train models for tasks like anomaly detection, from Mazak is the introduction of AI-assisted programming fea-
fault diagnosis, or process optimization. tures. These features enable their machines to learn from historical
• Image data: Images captured by cameras or sensors placed data, leveraging AI algorithms to generate optimized machining
within or around the CNC machine can be used to train mod- strategies. Mazak’s AI-enhanced systems also incorporate real-time
els for tasks such as object detection, defect identification, or process monitoring, allowing for the early detection of potential
quality control. These images can provide visual information issues. This proactive approach significantly reduces downtime and
about the workpiece, tooling, or any abnormalities during contributes to improved overall productivity. In addition, Mazak’s
the machining process. Mazatrol SmoothAi system stands out for its ability to generate
• Time series data: CNC machining operations often generate tool paths directly from the Mazatrol controller, requiring minimal
time-stamped data that captures the changes and dynamics human interaction, particularly when working with solid models.
over time. Examples include sensor measurements sampled Similarly, ESPRIT, a CNC machining software, is leveraging ar-
at regular intervals or historical process data. Time series tificial intelligence (AI) to enhance its capabilities [20]. ESPRIT
data can be used to train models for tasks like forecasting, utilizes AI algorithms to optimize machining processes and im-
process control, or pattern recognition. prove productivity in the manufacturing industry. The software
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Abdelaal et al.

employs machine learning to analyze and adapt to various machin- technical documentation and enable rapid prototyping and feature
ing scenarios, enabling it to make intelligent decisions and optimize refinement. Tools like Microsoft Copilot and Watson Code Assistant
tool paths. By leveraging AI, ESPRIT aims to reduce cycle times, utilize GenAI to facilitate faster and more efficient code develop-
enhance tool life, and improve surface finish quality. Additionally, ment, thereby reducing manual labor and improving deployment
there exist specific AI features in ESPRIT, such as Adaptive Spin- times.
dle Speed Control and AI-Driven Collision Avoidance, which help LLMs enable marketing teams to create rich, original content that
improve machining efficiency and prevent tool damage. resonates with the brand’s voice across multiple platforms. These
models personalize interactions, drive engagement, and improve
conversion by analyzing consumer data. They also provide insights
7 GENERATIVE AI FOR MANUFACTURING into market trends and consumer behavior and offer quick, accu-
In this section, we discuss the adoption of generative AI (GenAI), rate translations to help brands achieve global reach. Aside from
such as Large Language Models (LLMs) and Generative Adversarial marketing, in sales, LLMs automate interactions and personalize
Networks (GANs), to optimize the manufacturing process. Accord- communication This can drastically boost the efficiency of the sales
ing to a recent KPMG study, conducted in March 2024 with over process. LLMs help in lead generation, and content creation (such as
280 decision-makers, 53% of companies plan to increase their in- emails and proposals), and integrate with CRM systems to enhance
vestments in GenAI within the next 12 months, with half of these data management and follow-ups. Additionally, LLMs identify up-
companies aiming for a 40% or more increase [18]. Additionally, selling and cross-selling opportunities, assist in sales training, and
67% of companies anticipate a rise in turnover and automation, support advanced sales strategies like SEO-optimized product de-
65% expect reduced costs, and 55% hope for higher productivity. scriptions and automated contract management. LLMs transform
However, 37% of respondents fear potential negative effects, such customer service by powering chatbots and virtual assistants that
as job losses, data protection issues, and lack of employee accep- provide immediate, accurate, and personalized responses. They han-
tance. The study also highlights a need for improvement in strategy, dle multiple interactions across different platforms, enhancing the
governance, and training, as only about one-third of companies overall customer experience while freeing human agents to solve
have a strategy in place, only 8% have a comprehensive governance more complex issues.
model, and 63% do not feel adequately prepared for the EU’s AI Act. Aside from customer service, LLMs can be a handy tool for
Furthermore, only 38% of respondents believe their company is well product/service development. LLMs use extensive market research
or very well prepared to train employees in handling GenAI-based and consumer insights to aid in product and service development.
solutions. Makatura et al. [22] assessed the capabilities of LLMs across various
The manufacturing industry is currently grappling with numer- tasks, including design generation from natural language specifica-
ous challenges, including cost pressures, supply chain disruptions, tions; design space creation and variation; preparation of designs
and the need to continuously innovate while maintaining safety for manufacturing; design performance evaluation; and discovery
and stability. This sector is also under significant pressure to adopt of high-performing designs based on specific metrics. The authors
sustainable and eco-friendly practices amidst the global push for highlighted the ability of LLMs to generate designs from high-
energy transition and reduced carbon footprints. GenAI is emerging level textual input and handle diverse representations and problem
as a powerful solution to many of these challenges. As highlighted domains. LLMs demonstrate proficiency in creating parametric
in McKinsey’s 2023 AI report, GenAI offers various capabilities that designs, suggesting parameter bounds, and generating design vari-
are gaining traction in the manufacturing field [7]. For instance, ations within a given design space. They also discussed the ap-
summarization is a key application where GenAI helps to improve plications of LLMs in design for manufacturing (DfM), such as
the accessibility and usability of vast amounts of data stored in identifying suitable manufacturing techniques, suggesting design
documents such as manuals and logbooks. For instance, IBM de- modifications, and creating manufacturing instructions.
veloped GenAI accelerators specifically for manufacturing to aid Furthermore, they explored the potential of LLMs in evaluating
in this aspect, enhancing the efficiency with which operators and designs based on objective, semi-subjective, and subjective crite-
engineers access and utilize information [27]. LLMs can also im- ria. LLMs can assist in inverse design by generating cost functions
prove the management of complex documentation by efficiently and optimized designs based on constraints. In asset management,
organizing and parsing extensive texts, which facilitates quicker GenAI can develop foundation models for predictive maintenance,
data retrieval. They help in drafting, editing, and validating detailed especially when failure data is scarce. These models can be pre-
documents, ensuring accuracy and adherence to guidelines, which trained without labels and fine-tuned as needed, offering a robust
is crucial in technical and specialized fields. method for maintaining equipment and predicting failures. Despite
Manufacturing firms often struggle with siloed and inconsis- the potential of LLMs to streamline and enhance various aspects
tent data systems. GenAI can play a crucial role in creating a uni- of the design and manufacturing process, There exist several lim-
fied knowledge graph that provides contextual understanding and itations of LLMs in design and manufacturing. These limitations
accelerates data-driven decisions. IBM’s AI-powered Knowledge include challenges in analytical reasoning, complex computations,
Discovery system is an example of such an application, helping correctness, verification, and scalability. To address these limita-
to streamline data analysis and feature engineering in industrial tions, several potential solutions such as using domain-specific
settings. In software engineering, LLMs contribute to automated languages, leveraging APIs for complex computations, implement-
code generation, code review, and bug identification, which stream- ing automated verification, decomposing large tasks, and adopting
lines development processes. They assist in maintaining up-to-date incremental design approaches, can be investigated and adopted.
AI in Manufacturing: Market Analysis and Opportunities Conference acronym ’XX, June 03–05, 2018, Woodstock, NY

8 CONCLUSION interface-new-business-models-c0611749c8a5 Accessed: 2023-05-03.

[14] ifo Institute. 2023. Artificial Intelligence in Use at 13.3% of Companies in
In this arcticle, we thoroughly explored the integration and impact Germany. https://www.ifo.de/en/facts/2023-08-02/artificial-intelligence-use-
of AI within the German manufacturing sector, offering a market companies-germany Accessed: 2024-03-28.
[15] R. Kerris. 2021. Siemens Energy Taps NVIDIA to Develop Industrial Digital Twin
analysis supported by current data and forecasting future trends. of Power Plant in Omniverse. Nvidia Blog. https://blogs.nvidia.com/blog/2021/11/
We have seen how AI technologies, particularly in the realms of 15/siemens-energy-nvidia-industrial-digital-twin-power-plant-omniverse/ Ac-
predictive maintenance, process optimization, and safety enhance- cessed: 2023-04-19.
[16] S. Kinkartz. 2019. Germany lags on AI. https://www.dw.com/en/skeptical-
ments, are revolutionizing the industry, allowing for increased germany-lags-behind-on-artificial-intelligence/a-51828604
operational efficiency and a competitive edge in a rapidly evolving [17] E. Koptyug. 2019. Do you use artificial intelligence in your company, in the con-
market. The discussion highlighted the critical role of HMIs and text of Industry 4.0? https://www.statista.com/statistics/1243050/ai-industrial-
companies-industry-40-germany/#statisticContainer
presented a forward-looking perspective on the promising applica- [18] KPMG Germany. 2024. German Companies Plan Significant Investments in AI
tions of GenAI and LLMs, which are poised to further transform - KPMG Germany. https://kpmg.com/de/en/home/media/press-releases/2024/
05/german-companies-plan-substantial-investments-in-generative-ki.html Ac-
product design, customer support, and knowledge management cessed: 2024-04-03.
within manufacturing. Despite the challenges related to technology [19] E. A. Kumar. 2022. Milling Machine: Definition, Parts, Operation, Working Prin-
adoption, such as data quality and integration complexities, the ciple, Application, Advantages. https://themechanicalengineering.com/milling-
machine Available: https://themechanicalengineering.com/milling-machine.
potential for significant economic and operational benefits remains [20] A. Lancaster. 2021. Taking CNC programming to the next level with Artificial
high. Intelligence. https://amtil.com.au/esprit-cnc-machining-ai-amtil/ Available:
As we look to the future, it is clear that the continued exploration https://amtil.com.au/esprit-cnc-machining-ai-amtil/.
[21] Z. Lv and S. Xie. 2022. Artificial intelligence in the digital twins: State of the art,
and adoption of AI technologies will be crucial for the sustainability challenges, and future research topics. Digital Twin 12, 1 (2022), 12.
and growth of the manufacturing industry. The insights presented [22] Liane Makatura, Michael Foshey, Bohan Wang, Felix Hähnlein, Pingchuan Ma,
Bolei Deng, Megan Tjandrasuwita, Andrew Spielberg, Crystal Elaine Owens,
in this article should serve as a valuable resource for manufacturers, Peter Yichen Chen, Allan Zhao, Amy Zhu, Wil J. Norton, Edward Gu, Joshua
policymakers, and technology providers who are aiming to under- Jacob, Yifei Li, Adriana Schulz, and Wojciech Matusik. 2024. Large Language
stand and leverage AI/GenAI to innovate and thrive in the digital Models for Design and Manufacturing. An MIT Exploration of Generative AI (mar
27 2024). https://mit-genai.pubpub.org/pub/nmypmnhs.
age. The journey towards a fully AI-integrated manufacturing en- [23] M. Mohan Mewari and G. Kamath. 2021. 17 Remarkable Use Cases of AI in the
vironment is complex and challenging, but as demonstrated, it is Manufacturing Industry. Birlasoft. https://www.birlasoft.com/articles/17-use-
also rich with opportunities for transformation and advancement. cases-of-ai-in-manufacturing Accessed: 2023-04-18.
[24] Microchip. 2023. Smart Human Machine Interface: Complex Gesture Detection
at the Edge with Machine Learning. https://www.microchip.com/en-
REFERENCES us/solutions/machine-learning/smart-human-machine-interface# Available:
[1] 2024. Artificial Intelligence - Germany | Market Forecast. https://www.statista. https://www.microchip.com/en-us/solutions/machine-learning/smart-human-
com/outlook/tmo/artificial-intelligence/germany Accessed: 2024-03-14. machine-interface#.
[2] V. Agiang. 2021. How Intelligent Order Management Streamlines Fulfillment. [25] Gerd Mischler. 2019. KI für Industrie: Deutschland riskiert seine Führungsrolle.
Fabric Blog. https://fabric.inc/blog/intelligent-order-management/ Accessed: https://www.produktion.de/technik/zukunftstechnologien/kuenstliche-
2023-04-18. intelligenz/ki-fuer-industrie-deutschland-riskiert-seine-fuehrungsrolle-
[3] I. Aloyan. 2023. AI in CNC Programming: Revolutionizing the Manufacturing In- 335.html Accessed: 2024-01-19.
dustry with AI-Powered CNC Machines. https://mechutopia.com/the-revolution- [26] Renesas. 2023. HMI SoM with AI Accelerator. https://www.renesas.com/us/
of-artificial-intelligence-in-cnc-programming/ Available: https://mechutopia. en/application/communication-computing-infrastructure/high-performance-
com/the-revolution-of-artificial-intelligence-in-cnc-programming/. computing/hmi-som-ai-accelerator#overview Available: https://www.renesas.
[4] Avanada. 2023. Siemens AG Employees Get Instant Responses to Inquiries with com/us/en/application/communication-computing-infrastructure/high-
Chatbot. https://www.avanade.com/en/clients/siemens-chatbot Accessed: performance-computing/hmi-som-ai-accelerator#overview.
2023-04-18. [27] Sreejit Roy and Sanjay Prasad. 2024. Four Ways Generative AI Addresses Manu-
[5] Capgemini Research Institute. 2019. Scaling AI in Manufacturing Operations: facturing Challenges. IBM. https://www.ibm.com/blog/4-ways-generative-ai-
A Practitioners’ Perspective. https://www.capgemini.com/de-de/wp-content/ addresses-manufacturing-challenges/
uploads/sites/5/2019/12/Report-%E2%80%93-AI-in-MfG-Ops.pdf [28] Samsara. 2020. What is an HMI? https://www.samsara.com/guides/hmi/ Avail-
[6] Carotron. 2020. Human Machine Interface Guide. Carotron Articles. https: able: https://www.samsara.com/guides/hmi/.
//www.carotron.com/articles/hmi-benefits/ Accessed: 2023-05-03. [29] V. Soni. 2020. Edge AI-Modernizing HMI for Consumer and Smart Home
[7] Michael Chui, Lareina Yee, Bryce Hall, Alex Singla, and Alexander Devices. https://www.einfochips.com/blog/edge-ai-modernizing-hmi-for-
Sukharevsky. 2023. The state of AI in 2023: Generative AI’s breakout consumer-and-smart-home-devices/ Available: https://www.einfochips.com/
year. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the- blog/edge-ai-modernizing-hmi-for-consumer-and-smart-home-devices/.
state-of-ai-in-2023-generative-ais-breakout-year. Accessed: May 03, 2024. [30] M. Soori, B. Arezoo, and R. Dastres. 2023. Machine Learning and Artificial
[8] L. Cope. 2022. A Complete Guide to HMIs: What Are They, Types and Uses. En- Intelligence in CNC Machine Tools, A Review. Sustainable Manufacturing and
gineer Fix. https://engineerfix.com/electrical/hmis/what-is-a-hmi-and-where- Service Economics (2023), 100009.
are-they-used/ Accessed: 2023-05-03. [31] Striped Giraffe Team. 2021. More German companies see opportunities in artificial
[9] A. Dhapte. 2023. Global Human Machine Interface (HMI) Market Overview. Mar- intelligence. https://www.striped-giraffe.com/en/blog/more-german-companies-
ket Research Future. https://www.marketresearchfuture.com/reports/human- see-opportunities-in-artificial-intelligence/
machine-interface-market-1092 Available: https://www.marketresearchfuture. [32] Spatial Team. 2022. How Digital Twins Are Transforming Manufacturing. https://
com/reports/human-machine-interface-market-1092. blog.spatial.com/how-digital-twins-are-transforming-manufacturing Accessed:
[10] P. Diianni and W. DeGirloamo. 2021. BMW i Ventures Invests in Inventory 2023-04-19.
Optimization Software Company Verusen to Fuel Intelligent, Connected Supply [33] K. Tubbesing. 2019. Thyssenkrupp is developing its own AI. Hannover
Chains. BMW USA News. https://www.bmwusanews.com/newsrelease.do?id= Messe. https://www.hannovermesse.de/en/news/news-articles/thyssenkrupp-
3685 Accessed: 2023-04-18. is-developing-its-own-ai Accessed: 2023-04-18.
[11] N. Flaherty. 2021. Bosch builds digital twin for more accurate maps. Technology [34] P. Ulrich and V. Frank. 2021. Relevance and Adoption of AI technologies in
News. https://www.eenewseurope.com/en/bosch-builds-digital-twin-for-more- German SMEs – Results from Survey-Based Research. Procedia Computer Science
accurate-maps/ Accessed: 2023-04-19. 192 (2021), 2152–2159.
[12] M. Frąckiewicz. 2023. The Future of Human-Machine Interaction with AI and One- [35] T. von Hahn and C. Mechefske. 2022. Machine Learning in CNC Machining: Best
Shot Learning. https://ts2.space/en/the-future-of-human-machine-interaction- Practices. Machines 10, 1233 (2022).
with-ai-and-one-shot-learning/ Available: https://ts2.space/en/the-future-of- [36] Market Watch. 2023. Micro Milling Machine Market Report: A Comprehensive
human-machine-interaction-with-ai-and-one-shot-learning/. Analysis of Industry Growth with estimated future CAGR of 10.7% during
[13] A. Gonfalonieri. 2020. AI & Human-Machine Interface: New Business Models. 2023-2030. https://www.marketwatch.com/press-release/micro-milling-
Towards Data Science. https://towardsdatascience.com/ai-human-machine- machine-market-report-a-comprehensive-analysis-of-industry-growth-
Conference acronym ’XX, June 03–05, 2018, Woodstock, NY Abdelaal et al.

with-estimated-future-cagr-of-107-during-2023-2030-2023-04-26 Available: learning/research-on-ai-trends-in-manufacturing Accessed: 2023-04-12.

https://www.marketwatch.com/press-release/micro-milling-machine-market- [38] ZEW Research. 2024. Concerns and Uncertainty Impede AI Adoption in Com-
report-a-comprehensive-analysis-of-industry-growth-with-estimated-future- panies. https://www.zew.de/en/press/latest-press-releases/concerns-and-
cagr-of-107-during-2023-2030-2023-04-26. uncertainty-impede-ai-adoption-in-companies Accessed: 2024-05-21.
[37] D. Wee. 2021. New research reveals what’s needed for AI acceleration in manufac-
turing. Google Cloud Blog. https://cloud.google.com/blog/products/ai-machine-