Multi-Agent Control for Robotics

Multi-Agent Control for Robotics

Eytan Canzini

About the Project

For robots to have a meaningful impact in a variety of paradigms, they must move beyond single robot architectures. Multi-robot systems present a way to decentralise complex tasks - such as manipulation, exploration and sensing - to allow for more efficient task completion.

This research bridges multiple different application studies, with projects in manufacturing, agriculture and space systems. We are interested in methods that allow for decentralised perception, planning and control for multi-robot systems with a key emphasis on robustness and stability. 

Our robots include two UFACTORY xArm 6 manipulators and two ClearPath Dingo robots. When combined, these platforms allow us to perform cutting-edge research and develop novel algorithms to solve complex problems in both academia and industry.