Jonah Siekmann
Papers
2
Total Citations
14
H-Index
2
About
Jonah Siekmann is a leading researcher in legged robotics, specializing in sim-to-real reinforcement learning for bipedal locomotion. His work focuses on enabling robots to traverse complex, unstructured environments without relying on fragile, high-precision sensing systems. In his seminal 2021 paper, "Blind Bipedal Stair Traversal via Sim-to-Real Reinforcement Learning" (9 citations), Siekmann demonstrated that a bipedal robot could reliably climb and descend stairs using only proprioceptive feedback—a breakthrough that challenges the field's dependence on accurate terrain estimation. This work highlights his commitment to building robust, real-world systems that operate under uncertainty. Complementing this, his paper "Sim-to-Real Learning of All Common Bipedal Gaits via Periodic Reward Composition" (5 citations) introduced a novel framework for generating the full spectrum of bipedal gaits—walking, trotting, and running—through intuitive reward design. By composing periodic reward functions, Siekmann made it possible to transfer complex locomotion behaviors from simulation to reality with remarkable fidelity. His contributions are shaping the future of autonomous robots that can navigate human environments safely and efficiently, bridging the gap between simulation and the messy, unpredictable real world.
Research Focus
Key Achievements
Top Papers
- 1Blind Bipedal Stair Traversal via Sim-to-Real Reinforcement Learning9 citations · 2021
- 2