Learning robust autonomous navigation and locomotion for wheeled-legged robots
Problem
Wheeled-legged robots need adaptive locomotion over varied terrains and efficient navigation around complex dynamic obstacles in urban environments.
Why it matters
Autonomous urban robots could transform logistics efficiency, but failure on rough terrain or with obstacles is costly.
What the AI did
Model-free reinforcement learning with privileged learning trains a versatile locomotion controller, integrated hierarchically with a learned navigation controller and large-scale path planning.
Results
Controller achieves efficient, robust locomotion over various rough terrains with smooth transitions between walking and driving modes.
Impact
Presents a fully integrated learned system for robust wheeled-legged locomotion and navigation, per the abstract.
Record
| Venue | Science Robotics |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Joonho Lee, Marko Bjelonic, Alexander Reske, Lorenz Wellhausen, Takahiro Miki, Marco Hutter |
| Citations | 132 (OpenAlex, at import time) |
| References | 59 |
| Model or method | model-free reinforcement learning with privileged learning |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1126/scirobotics.adi9641 |
| Open access | green |
| Keywords | reinforcement learning, wheeled-legged robots, autonomous navigation, locomotion control, privileged learning |
Importance rating
significant — thousands to millions affected
Abstract states autonomous wheeled-legged robots could transform logistics systems and urban operational efficiency; validated by autonomous kilometer-scale navigation missions in Zurich and Seville.
More in Robotics & Autonomy
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Autonomous robotic laparoscopic surgery for intestinal anastomosis | 2022 | 384 |
7
|
| 2 | Solving Rubik's Cube with a Robot Hand | 2019 | 569 |
6
|
| 3 | PaLM-E: An Embodied Multimodal Language Model | 2023 | 346 |
6
|
| 4 | Assistive Navigation Using Deep Reinforcement Learning Guiding Robot With UWB/Voice Beacons and Semantic Feedbacks for Blind and Visually Impaired People | 2021 | 48 |
6
|
| 5 | Towards Deep Radar Perception for Autonomous Driving: Datasets, Methods, and Challenges | 2022 | 154 |
6
|