Education

This project is a follow-up to the course Software testing. It creates a “playground” where students can experiment with different testing techniques they learned during the Software Testing course. Moreover, testing is practised as just one phase in a more general, systems engineering process, where software meets hardware.The systems under test, a model railway and a salinity and temperature-controlled water plant, are inspired by real safety-critical systems, such as railway management systems and automatic insulin pumps. Both systems are controlled using an Arduino microcontroller and C++.

This course introduces the principles and methods of collective intelligence — how groups of simple, locally interacting agents produce robust, adaptive, intelligent group-level behavior. The course follows a single arc: the field map, the individual agent, the emergent collective, and engineering applications. Themes Six canonical domains are covered: collective motion (Vicsek, Reynolds boids), foraging and stigmergy (ant trails, ACO), division of labor (response thresholds), distributed decision-making (honeybee quorum, best-of-N), synchronization (Kuramoto), and opinion dynamics (DeGroot, Hegselmann–Krause). Concepts Across domains, students develop a unifying toolkit: the agent–environment–interaction framework, local rules with limited sensing, positive and negative feedback, phase transitions and order parameters, mean-field analysis, network topology, and robustness under perturbation. Research and practice Collective intelligence is an active interdisciplinary field at the intersection of computer science, statistical physics, biology, and the social sciences. The course connects these foundations to the engineered systems where they now operate — swarm robotics, federated and multi-agent machine learning, distributed consensus and blockchain protocols, multi-agent LLM systems, and crowd computing — and teaches students to recognize the collective-intelligence structure underlying them. Outcome Students leave with a working vocabulary for any collective system, hands-on experience implementing and analyzing an agent-based model, and the ability to formulate, measure, and communicate a collective-intelligence research question — transferable to research and to the design of modern decentralized systems.

In many application domains, AI systems must collect input from humans or collaborate with them to enhance their performance and decision-making processes. This course will introduce students in the AI program to the existing theories and methods for the design and evaluation of interactions between humans and [intelligent, socially interactive, or immersive] computing systems. The aim is to give students an understanding of the concepts of usability, user experience (UX), and user-centered design and to equip them with empirical research skills for prototyping and iterative testing of such interactive computing systems. In the practical part of the course, students will acquire hands-on research experience by collaborating in groups to design, conduct, and report on an HCI experiment with human participants.

The first half of the course will cover the basics of robotics and robot control. Topics will include forward/inverse kinematics, modeling of robots, Denavit-Hartenberg, the basics of dynamics, trajectory planning and optimization and control algorithms. In the second half, the course will give an introduction into robot learning and embodied AI, i.e., the combination of artificial intelligence & machine learning with robotics. This will include the discussion of model-free and model-learning approaches for robotic tasks, their disadvantages and advantages over control methods.

What does it mean when a robot steps out of your computer in physical space and starts interacting with humans as if it were a social entity itself? This course will address various aspects that come into play when physically present robots interact with humans. This course focuses mainly on social robots, humanoids in particular, suited for communicative purposes. We will discuss both sides of the coin: perspectives from the robot-side as well as perspectives from the user-side. Differences and similarities with common computer-based AI (cf. avatars, chatbots) are discussed in comparison to human-robot interaction, exemplified by main application areas (e.g., healthcare, education). Basic models and relevant aspects of the psychology of the user and ethical aspects are also discussed. If robots start talking to humans by themselves, how can they understand each other? Lectures will therefore also address the basics of the computational aspects involved (e.g., the Sense-Think-Act paradigm, and Large Language Models (LLMs)). How to translate theoretical models into effective computer models is another challenge that will not only be lectured on but also practised. A team of lecturers with the respective expertise will lecture on the wide variety of topics in this course. Several practical tutorials/workshops and a hackathon with real robots accompany the lectures to gain hands-on experience regarding the complexities of the human-centered human-robot interaction design and development process.

Pervasive (or ubiquitous) computing is a trend based on Mark Weiser’s vision of computers available “always and everywhere”, embedded in everyday life. This course is an introduction to pervasive computing systems that assist people in their daily lives. Think about a self-driving car, a speech-controlled wheelchair, an automatic insulin pump for diabetic patients, or a navigation system for visually impaired pedestrians. These systems:sense the context (time, user’s location, blood sugar concentration, road scenery, etc),recognize data patterns, reason and make intelligent decisions, andact upon the environment, by controlling the wheels or the pump motor, suggesting the best route, or just notifying a caretaker.The main components of such a system are sensors, controllers, and actuators. In this course, the students will learn different techniques to acquire signals from the environment, process these raw signals in order to infer context by using AI and in particular machine learning, and to write software agents for control. During the practical sessions, the students will get experience with these techniques and build their own microcontroller-based smart system. Programming is done in MATLAB and C++.