Bootstrapping Motor Skill Learning with Motion Planning
Not on the map. Why?
Abstract (arXiv)
Learning a robot motor skill from scratch is impractically slow; so much so that in practice, learning must be bootstrapped using a good skill policy obtained from human demonstration. However, relying on human demonstration necessarily degrades the autonomy of robots that must learn a wide variety of skills over their operational lifetimes. We propose using kinematic motion planning as a completely autonomous, sample efficient way to bootstrap motor skill learning for object manipulation. We demonstrate the use of motion planners to bootstrap motor skills in two complex object manipulation scenarios with different policy representations: opening a drawer with a dynamic movement primitive representation, and closing a microwave door with a deep neural network policy. We also show how our method can bootstrap a motor skill for the challenging dynamic task of learning to hit a ball off a tee, where a kinematic plan based on treating the scene as static is insufficient to solve the task, but sufficient to bootstrap a more dynamic policy. In all three cases, our method is competitive with human-demonstrated initialization, and significantly outperforms starting with a random policy. This approach enables robots to to efficiently and autonomously learn motor policies for dynamic tasks without human demonstration.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 20 of its 46 references that are in the map. First posted on arXiv 12 Jan 2021.
Builds on
20 of 46 references.
References · 20
- 2015Continuous control with deep reinforcement learningcited by 3,279
- 2016Generative Adversarial Imitation Learningcited by 1,576
- 2015End-to-End Training of Deep Visuomotor Policiescited by 847
- 2017Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrationscited by 685
- 2017Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewardscited by 302
- 2017Dex-Net 2.0: Deep Learning to Plan Robust Grasps with Synthetic Point Clouds and Analytic Grasp Metricscited by 259
- 2018Residual Reinforcement Learning for Robot Controlcited by 242
- 2019Category-Level Articulated Object Pose Estimationcited by 158
- 2018Residual Policy Learningcited by 146
- 2019A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithmscited by 136
- 2018Reinforcement and Imitation Learning for Diverse Visuomotor Skillscited by 124
- 2017Deeply AggreVaTeD: Differentiable Imitation Learning for Sequential Predictioncited by 93
8 more
- 2017Combining Self-Supervised Learning and Imitation for Vision-Based Rope Manipulationcited by 90
- 2016High precision grasp pose detection in dense cluttercited by 33
- 2019AC-Teach: A Bayesian Actor-Critic Method for Policy Learning with an Ensemble of Suboptimal Teacherscited by 27
- 2018Fast Policy Learning through Imitation and Reinforcementcited by 26
- 2016PLATO: Policy Learning using Adaptive Trajectory Optimizationcited by 23
- 2019Harnessing Reinforcement Learning for Neural Motion Planningcited by 19
- 2016Path Integral Guided Policy Searchcited by 14
- 2019Pixels to Plans: Learning Non-Prehensile Manipulation by Imitating a Plannercited by 4
Built on by
Citing papers per year: 2021: 1; 2022: 0; 2023: 0; 2024: 1.
2 papers.