ABSTRACT
Ubiquitous autonomous robots require novel methods for navigation and control that account for their interaction with other robots and humans.To enable autonomous operation, I will discuss constrained optimization methods for collision avoidance and formation control that apply to both ground and aerial vehicles. These methods model inter-agent cooperation, moving obstacles and the dynamic constraints of the robots. I’ll show experimental results with differential driven robots, wheelchairs, quadrotors navigating among humans and mobile manipulators collaboratively carrying objects. Towards human-in-the-loop multi-robot coordination, we have developed an interactive display formed by tens of small robots, and the method to specify high-level tasks for the multi-robot team and to automatically convert them into controllers with performance guarantees.