Efficient skill acquisition for insertion tasks in obstructed environments

Conference item

Data efficiency in robotic skill acquisition is crucial for operating robots in varied small-batch assembly settings. To operate in such environments, robots must have robust obstacle avoidance and versatile goal conditioning acquired from only a few simple demonstrations. Existing approaches, however, fall short of these requirements. Deep reinforcement learning (RL) enables a robot to learn complex manipulation tasks but is often limited to small task spaces in the real world due to sample inefficiency and safety concerns. Motion planning (MP) can generate collision-free paths in obstructed environments, but cannot solve complex manipulation tasks and requires goal states often specified by a user or object-specific pose estimator. In this work, we propose a robust system for efficient skill acquisition designed to address complex insertion tasks in obstructed environments. Our system leverages an object-centric generative model (OCGM) for versatile goal identification to specify a goal for MP combined with RL to solve complex manipulation tasks in obstructed environments. Particularly, OCGM enables one-shot target object identification and re-identification in new scenes, allowing MP to guide the robot to the target object while avoiding obstacles. This is combined with a skill transition network, which bridges the gap between terminal states of MP and feasible start states of a sample-efficient RL policy. The experiments demonstrate that our OCGM-based one-shot goal identification provides competitive accuracy to other baseline approaches and that our modular framework outperforms competitive baselines, including a state-of-the-art RL algorithm, by a significant margin for complex manipulation tasks in obstructed environments.

Authors: Yamada, J; Collins, J; Posner, I

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: PMLR
• Pages: 615-627
• Series: Proceedings of Machine Learning Research
• Series number: 242
• Publication date: 2024-06-11
• Acceptance date: 2024-04-01
• Event title: 6th Annual Learning for Dynamics & Control Conference (L4DC 2024)
• Event location: Oxford
• Event website: https://l4dc.web.ox.ac.uk/home
• Event start date: 2024-07-15
• Event end date: 2024-07-17
• EISSN: 2640-3498
• ISSN: 2640-3498
• Language: English
• Keywords: robotic manipulation; integrated planning and learning; reinforcement learning; motion planning; learning from demonstration