Learning agile and dynamic motor skills for legged robots
Explore this paper's citation graph
Summary
This work introduces a method for training a neural network policy in simulation and transferring it to a state-of-the-art legged system, thereby leveraging fast, automated, and cost-effective data generation schemes.
- Type
- article
- Published
- 2019-01-16
- Cited by
- 1,864
- References
- 67
- Access
- Open access
- OpenAlex
- https://openalex.org/W2911087563
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:58031572
Keywords
Reinforcement learning, Robot, Computer science, Agile software development, Legged robot
References
- Spring loaded inverted pendulum running: a plant model
- Open Dynamics Engine を用いたスノーボードロボットシミュレータの開発
- Acquisition of stand-up behavior by a real robot using hierarchical reinforcement learning
- PETMAN : A Humanoid Robot for Testing Chemical Protective Clothing (特集 2足歩行ロボット技術)
- Rigid Body Dynamics Algorithms
- Limit Cycles in a Passive Compass Gait Biped and Passivity-Mimicking Control Laws
- The biomechanics and motor control of human gait
- Variable-speed quadrupedal bounding using impulse planning: Untethered high-speed 3D Running of MIT Cheetah 2
- RHex: A Simple and Highly Mobile Hexapod Robot
- BigDog, the Rough-Terrain Quadruped Robot
- Design principles for highly efficient quadrupeds and implementation on the MIT Cheetah robot
- Resilient Machines Through Continuous Self-Modeling
- Model Learning with Local Gaussian Process Regression
- Hybrid position/force control of manipulators
- Discovery of complex behaviors through contact-invariant optimization
- Fast, robust quadruped locomotion over challenging terrain
- Actuator Dynamics and Delay Compensation Using Neurocontrollers
- Humanoid robot NAO: Review of control and motion exploration
- Learning Robot Dynamics for Computed Torque Control Using Local Gaussian Processes Regression
- A direct method for trajectory optimization of rigid bodies through contact
Cited by
- Nonsmooth Optimal Value and Policy Functions in Mechanical Systems Subject to Unilateral Constraints
- Data-Augmented Contact Model for Rigid Body Simulation
- A Survey on Policy Search Algorithms for Learning Robot Controllers in a Handful of Trials
- Learning to Walk via Deep Reinforcement Learning
- Robot learning—Beyond imitation
- Robust Recovery Controller for a Quadrupedal Robot using Deep Reinforcement Learning
- Sim-to-Real Transfer for Biped Locomotion
- Deep Visual MPC-Policy Learning for Navigation
- Robots on the run
- Iterative Reinforcement Learning Based Design of Dynamic Locomotion Skills for Cassie
- How to pick the domain randomization parameters for sim-to-real transfer of reinforcement learning policies?
- Habitat: A Platform for Embodied AI Research
- Can a Robot Become a Movie Director? Learning Artistic Principles for Aerial Cinematography
- Leg Trajectory Planning for Quadruped Robots with High-Speed Trot Gait
- From Video Game to Real Robot: The Transfer Between Action Spaces
- Stanford Doggo: An Open-Source, Quasi-Direct-Drive Quadruped
- Automated shapeshifting for function recovery in damaged robots
- Keep Rollin’—Whole-Body Motion Control and Planning for Wheeled Quadrupedal Robots
- Student-t policy in reinforcement learning to acquire global optimum of robot control
- PyRobot: An Open-source Robotics Framework for Research and Benchmarking
Related papers
- Development of Super-multi-legged Passive Dynamic Walking robot “Jenkka-III”
- Using Elastically Actuated Legged Robots in Rough Terrain: Experiments with DLR Quadruped bert
- Realization and motion analysis of multi-legged Passive Dynamic Walking
- Demonstration and Analysis of Quadrupedal Passive Dynamic Walking
- Machine Learning for Fast Quadrupedal Locomotion
- Technical Report on: Tripedal Dynamic Gaits for a Quadruped Robot
- Design, Simulation and Fabrication of Quadrupedal Robot Integrated using Five-Jointed Legs with Suspension Spring
- Enhancing Efficiency of Quadrupedal Locomotion Over Challenging Terrains with Extensible Feet
- Quadrupedal Locomotion: An Introduction to the Control of Four-legged Robots