Mingxin Yu

Mingxin Yu is a PhD candidate whose research is in the fields of robotics and machine learning. As a MathWorks Fellow, Mingxin will contribute to a research program aimed at developing a learning-based method for accelerating motion planning in robot manipulation problems and will help to develop a proposed, new approach that leverages a machine learning-based control barrier functions (CBF) in safety-critical motion planning scenarios to tackle the high sampling complexity of traditional sampling-based motion planning methods. The principal aim of this project is to incorporate CBF into the traditional motion planning methods to steer the system to navigate safely to the goal, while significantly reducing time-consuming sampling tasks and maintaining the guarantees of the original methods. Mingxin’s research has the potential to offer valuable new solutions to a planning problem with significant impact on numerous real-world applications, including manipulation in cluttered environments, multi-arm assembly tasks, and human-robot interaction.