Talk title: Dynamic cloth manipulation. Challenges in controlling cloth's motion
Abstract: Robot cloth manipulation has been attempted from several perspective over the last years. However, many previous approaches at manipulating cloths are done more from a planning perspective, than from a dynamics and control perspective. This results in slow motions and the robot taking too much time to, for example, fold a towel. To get closer to how us, humans, tackle cloth manipulation, dynamics needs to be taken into consideration. In this talk, I will discuss some of the approaches that we have taken to dynamically manipulate cloth or control the behaviors of some parts of a cloth garment. From Model-Free Reinforcement Learning, to Model-Based Deep Reinforcement Learning and Model Predictive Control.