Philipp Reist
Papers
8
Total Citations
282
H-Index
6
About
Philipp Reist is a robotics researcher whose work spans the frontiers of robot learning, dynamic manipulation, and simulation-driven design. His most impactful contributions lie in two key areas: massively parallel deep reinforcement learning for locomotion and the open-loop control of juggling robots. In his landmark 2021 paper, "Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learning" (over 100 citations), Reist demonstrated how to train locomotion policies on a single GPU in minutes—a breakthrough that dramatically accelerated real-world robot deployment. His earlier work on the "Blind Juggler" (2009–2012, 39+ citations) pioneered the open-loop stabilization of unconstrained balls, using clever paddle curvature and linear motor actuation to achieve stable juggling without any sensing. This elegant approach to dynamic manipulation earned recognition for its simplicity and robustness. More recently, Reist contributed to "Factory: Fast Contact for Robotic Assembly" (2022), addressing the long-standing challenge of simulating contact-rich assembly tasks. His research has shaped how roboticists think about combining simulation, parallelism, and minimal sensing to achieve complex behaviors—from walking to juggling to assembly.
Research Focus
Key Achievements
Top Papers
- 1
- 2Factory: Fast Contact for Robotic Assembly54 citations · 2022
- 3
- 4Design and Analysis of a Blind Juggling Robot39 citations · 2012
- 5
- 6Design of the Pendulum Juggler9 citations · 2011
- 7Factory: Fast Contact for Robotic Assembly3 citations · 2022
- 8Control of a swinging juggling robot2 citations · 2013