Multi-agent Collaborative AI


"Creating flexible coordination mechanisms for autonomous decision-making entities, allowing to adapt to changing environments, to interact flawlessly with humans, and to exchange privacy-sensitive data, in this way leveraging the power of AI in a highly connected and rapidly changing world."

In multi-actor systems, several computers interact with each other. They each make autonomous decisions. It is fundamental that none of the computers knows the entire system, nor has direct control over the other computers. This is interesting for self-driving cars that adapt to the behavior of people and other autonomous vehicles. Or for privacy-sensitive systems where one actor cannot make all his data known to the other actors. For example for applications in the banking world, the stock market or for network routing. The research focuses on the adaptivity, robustness, manageability and preconditions for the proper functioning of multi-actor systems.

Ann Nowé VUB AI Lab Management team
Leander Schietgat VUB AI Lab Management team + WP1 Lead: Use Cases
Bram Vanderborght VUB Robotics & Multibody Mechanics WP2 Lead: Multi-Agent Control Systems
Bart de Boer VUB AI Lab WP3 Hybrid Multi-Agent Systems for Collective Action
Jan Van den Bussche UHasselt WP4 Lead: Distributed Data Intelligence

Multiple research groups collaborate on this research domain. The above table only mentions the contact person and his/her affiliation.

More info...