Robotic Table Tennis: A Case Study into a High Speed Learning SystemWe present a deep-dive into a real-world robotic learning system that, inprevious work, was shown to be capable of hundreds of table tennis rallies witha human and has the ability to precisely return the ball to desired targets.This system puts together a highly optimized perception subsystem, a high-speedlow-latency robot controller, a simulation paradigm that can prevent damage inthe real world and also train policies for zero-shot transfer, and automatedreal world environment resets that enable autonomous training and evaluation onphysical robots. We complement a complete system deion, includingnumerous design decisions that are typically not widely disseminated, with acollection of studies that clarify the importance of mitigating various sourcesof latency, accounting for training and deployment distribution shifts,robustness of the perception system, sensitivity to policy hyper-parameters,and choice of action space. A video demonstrating the components of the systemand details of experimental results can be found athttps://youtu.be/uFcnWjB42I0.arxiv.org



- dc official App