The seminar is open to public and does not require any form of registration.
Over the last ten years, hybrid system theory has been increasingly used in the field of robotics, where the need for verification is often high, especially in contexts where robots interact with humans. However, up to now, most successful results were obtained with robots that can be considered as points, such as mobile robots, networks of unmanned vehicles, etc. Articulated robots (which are ubiquitous in some industries) present a different challenge due to their dynamical and geometrical complexity. The classical control framework for such robots can realize complex tasks with simple specifications, but it relies on numerical methods that leave a lot of uncertainty on the motions actually produced. I will introduce this framework and explain some of the specific issues associated with it. Then, I will present a few problems that examplify the type of difficulty that must be overcome to enable relevant synthesis and verification for articulated robots.