Achieving Human Level Competitive Robot Table Tennis
Citations Over TimeTop 10% of 2025 papers
Abstract
Achieving human-level performance on real world tasks is a north star for the robotics community. We present the first learned robot agent that reaches amateur humanlevel performance in competitive table tennis. Table tennis is a physically demanding sport that takes humans years to master. We contribute (1) a hierarchical and modular policy architecture consisting of (i) low level controllers with their skill descriptors that model their capabilities and (ii) a high level controller that chooses the low level skills, (2) techniques for enabling zero-shot sim-to-real and curriculum building, including an iterative approach (train in sim, deploy in real), and (3) real time adaptation to unseen opponents. Policy performance was assessed through 29 robot vs. human matches of which the robot won 45 % (13/29). All humans were unseen players and their skill level varied from beginner to tournament level. Whilst the robot lost all matches vs. the most advanced players it won 100 % matches vs. beginners and 55 % matches vs. intermediate players, demonstrating solidly amateur humanlevel performance. Videos of the matches can be viewed here1.See sites https://google.com/view/competitive-robot-table-tennis.
Related Papers
- → ARROCH: Augmented Reality for Robots Collaborating with a Human(2021)33 cited
- → “Should I Follow the Human, or Follow the Robot?” — Robots in Power Can Have More Influence Than Humans on Decision-Making(2023)18 cited
- → Modelling User Experience in Human-Robot Interactions(2015)10 cited
- → Human-Robot Interaction and Collaborative Manipulation with Multimodal Perception Interface for Human(2019)6 cited
- → Investigation of a Human’s Opinion Affected by Social Influence of a Group Norm in a Human-Robot Group After a Human-Robot Scenario(2020)