ROBEL: Robotics Benchmarks for Learning with Low-Cost Robots

Abstract

ROBEL is an open-source platform of cost-effective robots designed for reinforcement learning in the real world. ROBEL introduces two robots, each aimed to accelerate reinforcement learning research in different task domains: D'Claw is a three-fingered hand robot that facilitates learning dexterous manipulation tasks, and D'Kitty is a four-legged robot that facilitates learning agile legged locomotion tasks. These low-cost, modular robots are easy to maintain and are robust enough to sustain on-hardware reinforcement learning from scratch with over 14000 training hours registered on them to date. To leverage this platform, we propose an extensible set of continuous control benchmark tasks for each robot. These tasks feature dense and sparse task objectives, and additionally introduce score metrics as hardware-safety. We provide benchmark scores on an initial set of tasks using a variety of learning-based methods. Furthermore, we show that these results can be replicated across copies of the robots located in different institutions. Code, documentation, design files, detailed assembly instructions, final policies, baseline details, task videos, and all supplementary materials required to reproduce the results are available at www.roboticsbenchmarks.org.

Published @ CoRL2019. For details visit - http://www.roboticsbenchmarks.org

References in corpus (11)

• Continuous control with deep reinforcement learning
• Soft Actor-Critic Algorithms and Applications
• QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation
• DeepMind Control Suite
• Learning Dexterous In-Hand Manipulation
• Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Research
• Sim-to-Real Reinforcement Learning for Deformable Object Manipulation
• Stochastic Latent Actor-Critic: Deep Reinforcement Learning with a Latent Variable Model
• Active Domain Randomization
• REPLAB: A Reproducible Low-Cost Arm Benchmark Platform for Robotic Learning
• Dexterous Manipulation with Deep Reinforcement Learning: Efficient, General, and Low-Cost

Cited by in corpus (6)

• Efficient Exploration via State Marginal Matching
• TriFinger: An Open-Source Robot for Learning Dexterity
• Emergent Real-World Robotic Skills via Unsupervised Off-Policy Reinforcement Learning
• Interferobot: aligning an optical interferometer by a reinforcement learning agent
• DeepClaw: A Robotic Hardware Benchmarking Platform for Learning Object Manipulation
• Adaptation of Quadruped Robot Locomotion with Meta-Learning