Will Yang, a fifth year doctoral candidate, has been recognized for his work in dynamic motion robotics at the recent Robotics: Science and Systems (RSS) Conference which took place in Delft, Netherlands. Yang’s paper, “Dynamic On-Palm Manipulation via Controlled Sliding” received the Outstanding Student Paper Award.
The idea for this research originated from considering how to grab objects very quickly, where fully enclosing an object with a gripper would be too slow. Through collaborative refinement with lab mates, the project evolved to focus on utilizing sliding and friction in novel ways. “The paper represents a fairly significant step forward in our understanding of what types of problems can be solved with model-based control,” says Yang. “It’s one of the first papers to leverage sliding in a complex dynamic task.”
The paper explores the potential of leveraging sliding dynamics in complex robotic tasks, a concept that challenges traditional approaches to object handling in robotics. To grab an object quickly, a robot gripper would have been too slow. “The emphasis on utilizing sliding was actually unintentional, but when the robot doesn’t have a full rigid grasp on the object, sliding and friction become extremely relevant and we didn’t have many existing methods that could reason over those dynamics,” explains Yang. By demonstrating that sliding can be controlled rather than avoided, the research opens up new possibilities for recovering from unintentional sliding and enhancing overall robotic performance.
“I’m very happy that my paper was recognized for its contributions,” Yang comments. “I’m really excited about the results and am glad that this award is bringing more awareness to the work.” As he prepares to graduate, Yang will be joining Amazon Robotics as an Applied Scientist, where he plans to continue his work on dynamic object manipulation in warehouse robotics.
Yang is advised by Assistant Professor Michael Posa and works in the Dynamic Autonomy and Intelligent Robotics (DAIR) Lab.