The MSc program in Artificial Intelligence in Industrial Production offers a comprehensive curriculum that covers a wide range of cutting-edge topics in the field of AI for industrial production. With a focus on real-world applications and hands-on experience, students will gain a deep understanding of the latest trends and advancements in this exciting field. From machine learning and data analytics to the design of digital twins, autonomous mobile robots, and more, the following list of courses provides a glimpse into the program’s curriculum and the comprehensive education that students will receive. The structure of our MSc program can be seen here.

Major course units

Machine Learning and Data Analytics in Industrial Production

This course covers a wide range of topics in the field of machine learning, including supervised and unsupervised learning, deep learning, and reinforcement learning. It also focuses on the practical applications of machine learning and data analytics in industry, with a specific emphasis on cloud AI. The course is designed to equip students with the skills and knowledge necessary to analyze complex data sets, make predictions, and use these insights to drive positive change in the industrial production sector. Read more.

Justification: This discipline is included in this master program to facilitate the analysis of data obtained from the various equipment and operations in the production processes, data collected in various ways and stored in the cloud. The technologies necessary for data collection and preparation are taken into account, as well as those that help with their analysis (correlations, regressions, classifications, groupings). Other forms of artificial intelligence, from the category of neuro-symbolic models, are also applied in this field, especially for operating with small data sets and for assisting in decision-making. Knowing what data to collect for each process is essential, as is the most appropriate machine or deep learning models for each process, and how to interpret the results. Also, the process of adopting artificial intelligence in the enterprise (step-by-step) must be known. Creating the ability to optimize processes by uncovering less visible aspects and superior understanding of process capability and maturity are expected outcomes.

Design of Digital Twins

The course is designed to provide students with a comprehensive understanding of the creation and implementation of digital replicas of real-world systems. The focus is on learning how to use digital twins to monitor and optimize physical systems, improving their performance, and making data-driven decisions. This course also covers the use of cloud AI in digital twin design, enabling students to understand the benefits of using the cloud to enhance the functionality and scalability of digital twins. The curriculum includes hands-on projects and case studies to give students practical experience in the design and implementation of digital twins. Read more.

Justification: A digital twin is a digital representation of an intended or actual physical product, system or process that serves as its indistinguishable digital counterpart for practical purposes such as simulation, integration, testing, monitoring and maintenance. Digital twins are essential to contextualize the collected data. They are based on I-IoT, virtual models (including 3D graphic models), simulations. Machine learning models are extremely useful in relation to predictive maintenance. But without understanding which data must be collected at the level of equipment and technological installations specific to industrial production, how they can be collected and under what conditions, it is impossible to apply artificial intelligence models for various analyses. The digital twin is a central element in concepts that deal with the ever-increasing challenges of shorter production cycles, individualized products, massive cost pressure and shorter development times. In order to develop and execute in these framework conditions high-quality products with adequate sustainability and efficiency, the traditional “trial-and-error” method of physical construction of models and prototypes is still valid only conditionally. On the digital twin, new ideas, concepts and innovations can be tested through simulations and virtual commissioning. If the digital twin is correctly composed in unitary concepts and continuously updated, not only simulation in the preliminary phase, but also controlled intervention during the production regime is possible. This “real-time capability” opens up new possibilities for use in smart factories. In addition to time and cost savings, in particular the flexibility of the production facilities as well as the load capacity of the facilities are in the foreground.

Autonomous Mobile Robots

This course is focused on exploring the implementation of autonomous mobile robots in industrial production settings. Students will gain hands-on experience in programming and operating these robots, as well as understanding the various challenges and limitations involved in deploying them in real-world environments. The course will delve into topics such as robot localization and navigation, perception, mapping, and control systems, as well as how these technologies can be integrated and optimized for use in industrial applications. With a strong emphasis on practical skills and real-world examples, students will gain a deep understanding of how to design, develop, and deploy autonomous mobile robots to drive innovation and improve efficiency in industrial production. Read more.

Justification: Autonomous mobile robots are present in industrial enterprises for logistics operations. Autonomous navigation in dynamic environments involves both a specific sensory architecture and navigation and trajectory planning algorithms. Data comes from integrated laser scanners, 3D cameras, accelerometers, gyroscopes, radar systems and more to help generate the most effective decisions for each situation. The artificial intelligence algorithms encountered in such systems are from the area of automatic learning, reinforced learning, fuzzy logic, but also those specific to the automatic mapping of the environment, etc. Aspects of optimizing the design of these robots using evolutionary algorithms are also needed, as well as agent-based modeling. Deep learning models (including based on neural networks) are also used for environment recognition. In factories based on Industry 4.0 technologies, mobile autonomous robots are key components, including for the application of advanced production concepts such as agile and lean production.

Collaborative Robotic Systems

This course provides a comprehensive education on the implementation of robotic systems that work in collaboration with humans. This course focuses on the latest technologies and best practices for integrating robotic systems into various industrial settings, such as manufacturing, and logistics. Students will learn about the key concepts and techniques for creating effective human-robot collaboration systems, including motion planning, control, and human-robot interaction. Through hands-on projects and real-world applications, students will gain a deep understanding of the design, programming, and deployment of collaborative robotic systems, preparing them for careers in this exciting field. Read more.

Justification: A cobot or collaborative robot is a robot intended for direct human-robot interaction in a shared space or where humans and robots are in close proximity. Cobots are still industrial robots in fact, but with certain design specificities and replace traditional industrial robots where human-robot contact is required during production. Artificial intelligence models are applied in many cases specific to cobotic applications, from multimodal human-robot interfaces, to optimizing the behavior of robot joints in contact with the operator, to optimizing human-robot synchronous operations or to intervention in unstructured spaces.

Cognitive and Social Robotics

The course focuses on the development of cognitive capabilities in social robots that can interact and cooperate with humans and other robots in various settings. Students will learn about the cognitive and social capabilities required for robots to function effectively in human environments, as well as the latest technologies and algorithms for achieving these capabilities. Topics covered include natural language processing, computer vision, decision-making, and social signal processing. With a hands-on approach, students will have the opportunity to apply their knowledge to real-world problems and gain practical experience in developing cognitive and social robots. Read more.

Justification: Cognitive robotics is a subfield of robotics concerned with endowing a robot with intelligent behavior by providing it with a processing architecture that will enable it to learn and reason about how to behave in response to complex goals in a complex world. Social robotics is a subfield in cognitive robotics, applicable to use cases where human-robot interaction is deep. Currently, as a result of developments in personalized manufacturing, but also in the context of the development of resilient production systems, industrial robotics systems (especially collaborative ones, but not only) integrate components from the area of social robotics (see the production of Tesla, Toyota , etc.). This category also includes virtual assistants (robot-kiosks or chatbots based on NLP and QAM algorithms) to assist operators in small or customized series production. From the perspective of this discipline, the understanding of the various AI technologies incorporated in the architecture of a social robotic system will be approached, up to the level of communication with hardware elements, specific and general languages for the creation of human-robot dialogue, the way of integrating specialized libraries in this sense (including from the NLP area), etc., but also forms of interaction in the case of human-robot remote collaboration.

Lean and Agile Production

The course is focused on teaching the principles and practices of Lean and Agile methodologies for the optimization of industrial production processes. The course covers topics such as value stream mapping, inventory management, continuous improvement, and Lean product and process development. Students will learn how to apply Lean and Agile concepts to real-world production scenarios and how to effectively lead and implement change in a production environment. The objective is to equip participants with the skills necessary to increase efficiency, reduce waste, and improve overall production performance. Read more.

Justification: Lean manufacturing:. In the supply chain and production, lean principles analyze the steps in detail to understand where there is still non-value in order to reduce or eliminate it. In lean manufacturing all contributors to the production chain must operate “lean and mean” for the process to function properly. Thus, data on the operation of each operation (times, process parameters, etc.) are required. Agile manufacturing: While lean manufacturing focuses on eliminating areas of no value, agile manufacturing focuses more on the intelligent use of current resources and ensuring that the organization has the right data to implement changes in production. Accurate data is vital to making changes in production. By being agile or flexible, a company has the potential to adjust its manufacturing process – but more importantly and applicable to logistics, making changes based on evolving changes in the industry. To optimize agile / flexible production, the collected data can be analyzed with specific artificial intelligence models, especially evolutionary algorithms, but also machine learning models. The major objective pursued is to prepare the conditions for the generation of data on the streams of added value creation in the primary and secondary processes.

3D Printing Technologies

This course will focus on exploring the advancements and applications of 3D printing in various industries. The goal of this course is to provide students with a comprehensive understanding of the capabilities and limitations of 3D printing technology and how it can be utilized in various industries for efficient and cost-effective production. Students will learn about different types of 3D printing technologies, including Fused Deposition Modeling (FDM), Stereolithography (SLA), and Selective Laser Sintering (SLS). The course will also cover the materials used in 3D printing and their properties, as well as design considerations for 3D printing and post-processing techniques. Students will also learn how AI can enhance the design, production, and optimization of 3D printed objects. In this area, topics include AI-powered design tools, cloud-based simulation, and machine learning algorithms for material selection and print optimization. The course will provide hands-on experience with state-of-the-art 3D printing software and hardware, and students will develop a deep understanding of how AI can revolutionize the field of 3D printing. Read more.

Justification: 3D printing technologies help generate and collect data for the optimal design of mechanical products. Moreover, by combining 3D printing with generative design algorithms, it is possible to design innovative products and components, optimized in relation to the various materials used and loads to be supported. Machine learning models can be used to monitor and adjust the 3D printing process to correct errors in real time, considering the use of various innovative materials in the construction of products or prototypes. Without a deep understanding of the additive manufacturing technology process and the specifications of the materials used, developing machine learning models are useless. In the 3D printing process, artificial intelligence helps maintain the material properties of complex alloys such as titanium, carbon and other metals. The resulting models can be used for predictive maintenance. Machine learning even helps manufacturers improve spare parts and predict when to replace them. AI has other advantages when it comes to 3D printing, including the ability to analyze an object before the process begins and predict part quality. The use of machine learning algorithms also improves the fastening process and reduces waste in manufacturing. Machine learning in 3D printing is also applied to automatic monitoring of 3D printed parts. By integrating image processing and camera data, machine learning algorithms can detect defects during the printing process and help repair them without human intervention. AI technology can also help reduce the cost of reprinting parts. Another economic application of artificial intelligence in 3D printing is the guarantee of high-precision prints and quality of execution. The application of modern technologies for the rapid prototyping of mechanical components and products is an objective pursued by this discipline. By automating processes, AI can help eliminate human error and improve 3D printing performance. The fusion of AI and 3D printing has a number of benefits for manufacturing and quality control. It also helps accelerate the rise of Industry 4.0 and the Industrial Internet of Things.

3D Scanning Systems

The course focuses on 3D scanning technologies and the integration of artificial intelligence into these technologies. Students will learn about the various types of 3D scanning systems, including laser scanning and structured light scanning, and how they are used to capture and create digital representations of physical objects. They will also explore how AI algorithms can be applied to 3D scanning data to enhance the accuracy and efficiency of the scanning process, as well as to analyze and manipulate the resulting 3D models. Read more.

Justification: 3D scanning systems are systems intended for the collection of massive data associated with various objects, in order to apply (reverse) reconstruction engineering. 3D laser scanning is also used to measure vibrations in dynamic systems in industrial production. Once this massive data is collected, machine learning models are used for preventive maintenance. 3D scanning is also used to collect data in quality control, which can then be used in artificial intelligence models for analytics (typically machine learning models). 3D robotic scanning can be deployed in in-line or near-line factories to collect high-resolution 3D measurement data. 3D data contains a depth of information beyond what 2D vision sensors capture. There are algorithms capable of reconstructing a 3D map from multiple 2D images, but the resulting 3D map tends to be less accurate.

Virtual Reality in Production Systems

The course is focused on exploring the use of VR technology in industrial production. Students will learn how VR can be used to enhance production processes, including prototyping, training, and design visualization. They will also study the potential of VR to improve collaboration and communication within production teams. The course will cover the technical aspects of VR and its integration with existing production systems, as well as the development of VR applications and solutions. Students will gain hands-on experience in using VR tools and software to design and implement virtual production environments. Additionally, the course will address the challenges and opportunities associated with the adoption of VR in industrial production and its impact on the future of work. Read more.

Justification:  Virtual reality in the context of this study program has a strong connection with the design of digital twins. Virtual reality systems are augmented with sensors to retrieve data or extract information from the virtual environment based on operators’ reactions. Among other things, without this data, systems used in industrial production such as exoskeletons or the ergonomic design of systems would be much more difficult to achieve.

Knowledge Bases in Design

The course is focused on the use of knowledge bases and expert systems in the field of product design. It covers topics such as the creation, representation and maintenance of design knowledge, as well as the use of these systems to support decision-making and improve the design process. The course also explores the application of artificial intelligence and machine learning techniques to the development of intelligent design systems. Students will learn about the benefits and limitations of using knowledge bases in design and will develop practical skills in the use of knowledge-based systems through hands-on projects and case studies. Read more.

Justification: KBE (knowledge-base engineering) systems intend to capture product and process information to enable companies to replicate or model engineering processes. The model will then be used to automate parts of the process or all of it. The product model serves as a computer representation of the design process. In addition, models can contain physics-based analysis, databases, spreadsheets, legacy programs, and cost models, essentially other information outside of its environment. In this discipline students will understand how to create knowledge from various sources, how to validate it, how to represent it, how to generate knowledge-based inferences and how to create the ability to explain and justify the use of knowledge for various contexts. Students will come into contact with various software tools for engineering and knowledge base management in the context of product design (eg Protégé, PLTool, Prodigy, myPDDL).

Generative Design and Topological Optimization

The course is a cutting-edge educational program that focuses on the use of advanced algorithms and artificial intelligence to design more efficient and sustainable products and structures. In this course, students will learn about generative design, a process in which algorithms generate multiple design options based on input parameters, and topological optimization, which uses mathematical models to identify the optimal design from the generated options. The course will also explore the integration of AI in product design and the impact it has on sustainability, manufacturing, and the overall production process. The program provides hands-on training and real-world applications, allowing students to gain practical experience in the use of AI in the design process. Read more.

Justification: Generative design is a trend in optimal product design. Generative design is an iterative design process that involves a program that will generate a certain number of outputs that satisfy certain constraints and a designer that will fine-tune the feasible region by selecting a specific output or changing input values, ranges, and distribution. Evolutionary algorithms are applied in the composition of CAD systems for generative design. Specialists must know the process of applying these algorithms in constructive design and quantitative optimization. Generative design uses artificial intelligence and machine learning to transform tedious engineering design processes into a sophisticated yet natural interaction between computer and engineer. The main part of topology optimization and simulation is handled automatically by the computing unit. Additionally, feedback loops are shortened by lowering barriers to design. As a result, the engineer has more room to tackle challenges that require “common sense” or cannot be solved by computational systems.

Digitalization of Supply Chains

This course focuses on the digitalization of supply chains, exploring how new technologies, such as artificial intelligence and the Internet of Things, are revolutionizing the way goods and services are delivered to consumers. The course covers the key concepts and strategies for modernizing supply chains, including advanced data analytics, real-time tracking and monitoring, and the integration of digital technologies across the entire supply chain network. Through a combination of lectures, case studies, and hands-on exercises, students will gain a deeper understanding of how to design, implement, and manage digital supply chains that are efficient, secure, and sustainable. Read more.

Justification: Customer-supplier chains (internal and inter-company) are essential in value chains. The collection of data from all these processes allows the optimization of business systems through the use of artificial intelligence models, especially models from the category of automatic learning, but also those from the category of evolutionary algorithms, fuzzy logic, and those based on agents. The major objective pursued is the creation of capabilities on internal and external supply chains towards connected factories in preparing conditions for data generation on value-added streams in primary processes.

Green and Digital Transformation

This course explores how the integration of AI and other digital technologies can drive sustainable and efficient production processes while reducing environmental impact. The course covers topics such as energy-saving techniques, eco-friendly production methods, and the use of data analytics to optimize production processes, as well as the social and ethical implications of the transition to green and digital production. Students will learn how to analyze and design sustainable production systems, and how to evaluate the potential impact of different digital solutions on both the environment and the bottom line. Read more.

Justification: The aim is to prepare an industrial enterprise to know how to adapt its business models and organizational culture in order to use digital technology and digitization to create new directions of added value and meet the requirements of circular economy and orientation on new economic models driven by resilience, life cycle, servitization, shared resource use, etc.). A major objective pursued is to create the conditions to allow a company in the industrial zone to reach the level where the adoption of AI can bring added value.

Security of Industrial Networks

This course focuses on the security of industrial networks, including the protection of communication systems and data transmission in industrial environments. The course covers topics such as network architecture and protocols, threat analysis and risk assessment, as well as various methods for securing industrial networks, such as firewalls, access control, and encryption. The course also explores the integration of AI and machine learning techniques for real-time monitoring and threat detection in industrial networks. Upon completion, students will have a comprehensive understanding of the challenges and solutions for securing industrial networks and the role of AI in enhancing network security. Read more.

Justification: This discipline covers user-level topics such as: switching and monitoring; address allocation in the industrial network; routing and firewall functions; VLAN-separated production networks; Network Address Translation (NAT); Virtual Private Networks (VPNs); encrypted communication; authorization and access control; scalability through group and role concepts; how a PKI works; management of digital certificates. Machine learning models are used to improve security performance. Specialists also need to prepare themselves in understanding how to train models based on neural networks to deal with adversarial attacks that endanger the security of industrial networks, especially I-IoT.

Augmented Reality in Production Systems

This course focuses on the integration of Augmented Reality (AR) technology in industrial production systems. Participants will learn about the benefits of AR in enhancing productivity and quality, as well as its potential for creating a safer and more efficient production environment. Topics covered include the latest AR tools and applications, how to design and implement AR systems, and the role of AI in optimizing AR technology. By the end of the course, participants will have a solid understanding of AR in industrial production and its impact on the future of manufacturing. Read more.

Justification: Augmented reality (AR) is the process by which it is possible to superimpose a virtual reality on top of the real, concrete world observed with the naked eye. The applications of this technology can be found in many fields of activity, including in the field of industrial production, to assist operators in maintenance operations, in the use of equipment, in inspection operations, in assembly operations, etc. The driver for the adoption of AR in industrial manufacturing is the need for operators capable of agile operation in complex workloads. An industry where neglecting a small detail can be fatal (eg aircraft manufacturing), with AR an amazing 96% reduction in inspection time for already built components can be achieved. Inspection time is reduced from 36 hours to just 90 minutes. Of course, this is an extreme example, but a drop of at least 25% is typical for industrial production. In order to realize AR solutions in the industrial environment, there is a need for the development of complex, detailed 3D models of physical objects, there is a need for 3D simulation of the functioning of the systems and their integration in the AR technology architecture. In the context of this discipline, students will learn to be users of platforms that bring AI to AR for a range of benefits. For example, they will be able to use canonical AR SDKs (eg ARKit and ARCore) to place and manipulate objects in scenes, combine data from a device’s sensors to build the 3D world, track motion, render objects digital and to mediate interactions between digital and physical content. Also, Core ML and TensorFlow Lite are AI frameworks for mobile devices that provide control over the input and output of models and allow developers to input their own custom models, which are then trained to perform specific tasks for AR applications. The most common way to combine AR and AI models is to take images or sounds from a scene, run that data through a model, and use the model’s output to trigger effects within the scene. AI models are applied within AR and for use cases such as sound recognition, text recognition and translation, dynamic object position and orientation estimation, semantic segmentation and occlusion, object detection, scene or image labeling, etc.

Integrated Project on Artificial Intelligence in Industrial Production (4 semesters)

The project for the MSc program in Artificial Intelligence in Industrial Production is designed to be a comprehensive exploration of the application of AI technologies in industrial production processes. It is designed to be a 4-semester program that provides students with an in-depth understanding of how AI can be leveraged to improve productivity, efficiency, and competitiveness in manufacturing and other industrial processes. Students will have the opportunity to work on real-world projects, using state-of-the-art tools and technologies to develop and test AI-powered solutions to real industrial problems. Read mode.

Justification: The relevance of this project is based on the growing demand for advanced technologies in industrial production and the increasing importance of Artificial Intelligence (AI) in optimizing processes and enhancing efficiency. By undertaking a 4-semester project in AI in industrial production as part of a Master program, students will be able to acquire the necessary skills and knowledge to effectively apply AI in real-world industrial settings. The project would involve researching and developing AI models and applications that can improve various aspects of industrial production, such as predictive maintenance, quality control, energy management, and supply chain optimization. Additionally, the project would involve exploring the challenges and limitations of AI in industrial production and developing strategies to overcome these challenges.