Robotic Table Tennis: A Case Study into a High Speed Learning System

We present a deep-dive into a real-world robotic learning system that, in previous work, was shown to be capable of hundreds of table tennis rallies with a human and has the ability to precisely return the ball to desired targets. This system puts together a highly optimized perception subsystem, a high-speed low-latency robot controller, a simulation paradigm that can prevent damage in the real world and also train policies for zero-shot transfer, and automated real world environment resets that enable autonomous training and evaluation on physical robots. We complement a complete system description, including numerous design decisions that are typically not widely disseminated, with a collection of studies that clarify the importance of mitigating various sources of 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 system and details of experimental results can be found atthis https URL.

@article{dámbrosio2025_2309.03315,
  title={ Robotic Table Tennis: A Case Study into a High Speed Learning System },
  author={ David B. DÁmbrosio and Jonathan Abelian and Saminda Abeyruwan and Michael Ahn and Alex Bewley and Justin Boyd and Krzysztof Choromanski and Omar Cortes and Erwin Coumans and Tianli Ding and Wenbo Gao and Laura Graesser and Atil Iscen and Navdeep Jaitly and Deepali Jain and Juhana Kangaspunta and Satoshi Kataoka and Gus Kouretas and Yuheng Kuang and Nevena Lazic and Corey Lynch and Reza Mahjourian and Sherry Q. Moore and Thinh Nguyen and Ken Oslund and Barney J Reed and Krista Reymann and Pannag R. Sanketi and Anish Shankar and Pierre Sermanet and Vikas Sindhwani and Avi Singh and Vincent Vanhoucke and Grace Vesom and Peng Xu },
  journal={arXiv preprint arXiv:2309.03315},
  year={ 2025 }
}