In environments that become smarter every day, robots will increasingly share this space with humans and other robots while executing their tasks autonomously. For example, intelligent vehicles will provide mobility on-demand and deliver packages and autonomous boats and drones will perform environmental monitoring. These tasks require them to move in a complex, uncertain and changing environments. The challenge for these robots is to behave in such a way that it is safe and socially intuitive. The researcher aims to develop novel control and coordination methods that grant high performance and demonstrate safe motion through changing, dynamic and crowded environments.

The researcher is looking at two scenarios for control and coordination. The first one is when robots and humans can explicitly communicate and coordination is achieved via distributed constrained optimization. Since not all agents will explicitly communicate with the robot, think of a biker or a pedestrian, the researcher is also focusing on methods for communication-less implicit coordination based on intentions and movement. Here the researcher employs machine learning to model the interaction and provide guidance for a motion planner, which  then computes a trajectory for the robot that is safe, satisfies the task and respect its dynamics.