AI for Robotics Lab
The AI for Robotics Lab (AIR-Lab) in the MindLabs building provides space and equipment to study robotics and artificial intelligence.
The AIR-Lab is located on the ground floor of the MindLabs building and was set up with funding by the Municipality of Tilburg. The lab is managed by Tilburg University’s Department of Cognitive Science and Artificial Intelligence.
The types of experiments that can be conducted in the lab are:
- Fundamental research in robotics
- Human-robot interaction
Facilities
The lab equipment includes the following:
- Franka FR3 robot arm;
- Unitree GO1 robot quadruped;
- In-house built 16 degrees-of-freedom Tilburg hand;
- NAO humanoid robots;
- Pepper humanoid robot;
- QT robot;
- Furhat robot;
In addition to the above robot platforms, the lab also has basic electronic equipment.
Research projects
The AIR-lab focuses on recent advances in AI for robotics to endow robots with advanced cognitive capabilities and state-of-the-art performance. Research is organized along two main areas:
- CoDeRs: The Cognitive Developmental Robotics (CoDeRs) approach integrates insights from various fields - including machine learning, interaction design, computational neuroscience, cognitive/developmental psychology, etc. - to conduct interdisciplinary research in robotics and artificial intelligence. Goal is to develop computational models for embodied robots, both physical and virtual, that can interact with humans and other robots.
- DL4Robotics: The DL4Robotics line focuses on incorporating the latest advances in deep learning into robotics. Particular interest lies in the development of fully end-to-end robot learning methods based on deep reinforcement learning, imitation learning, and self-supervised learning. Applications span a variety of domains, including human-robot interaction and dexterous manipulation (manipulating objects by using precisely controlled forces and movements).
Some of the research currently taking place:
Large non-verbal Language Models (LnvLM)
When we think of communication, our first thought is language. Following this intuition, a vast amount of research in robotics (in particular, human-robot interaction (HRI)) has been dedicated to modeling spoken language. Nonetheless, a significant portion of information exchanged in human-human communication is non-verbal. Despite its importance, current non-verbal behavior by robots in HRI settings is either manually pre-programmed or controlled by humans via teleoperation. The objective of the project is to build on recent advances in language modeling and artificial intelligence to create large language models for non-verbal communication, including gestures, eye gaze, facial expressions, proxemics, etc. The developed models will be deployed on various robot platforms to test the ecological validity of the proposed approach. Part of the validation will involve experiments with Reinforcement Learning from Human Feedback (RLHF), relying on the trained embeddings of non-verbal communication to both specify tasks and extract reward signals.
Tilburg-CoDeRs: RoboCup Standard Platform League
AIR-Lab hosts a RoboCup team for the Standard Platform League, a robot soccer league, in which all teams compete with identical robots. The team is composed of students who conduct research on the perception, locomotion, and coordination of multiple humanoid robots in a competitive setting.
Intended users
The AI for Robotics Lab can be used by Tilburg University staff and students and by partners associated with MindLabs. The lab can only be used with the permission of the lab coordinator. In some cases, the presence of an authorized supervisor is also required.
Please note that most of the robot and hardware platforms are fragile and may require considerable prior experience to use. As such, and because the primary focus of the lab is on research, access is prioritized to users with previous experience in robotics.
Contact
If you are interested in using the lab and would like to discuss the options, please contact lab coordinators Murat Kirtay and Giacomo Spigler.
