URPlanner: A Universal Paradigm for Collision-Free Robotic Motion Planning Based on Deep Reinforcement Learning
Fengkang Ying, Hanwen Zhang, Haozhe Wang, Huishi Huang, Marcelo H Ang Jr
AI summary
Problem
Existing DRL-based motion planners for redundant manipulators are computationally costly, overly dependent on minimum distance calculations, and suffer from poor exploration and inefficient data utilization.
Approach
The method parameterizes the environment using bounding boxes and line segments to create a distance-independent obstacle avoidance reward, combined with an enhanced exploration algorithm and a data diffusion strategy to train policies from minimal expert demonstrations.
Key results
- Universal obstacle avoidance reward independent of minimum distance
- Augmented policy exploration and evaluation algorithm for stable DRL training
- Expert data diffusion strategy generating large datasets from few demonstrations
- Platform-agnostic planning that bypasses inverse kinematics for arbitrary manipulators
Why it matters
Provides a scalable, deployment-ready framework for training collision-free motion policies on any redundant robot without costly simulations or custom inverse kinematics.
Abstract
Collision-free motion planning for redundant robot manipulators in complex environments is yet to be explored. Although recent advancements at the intersection of deep re- inforcement learning (DRL) and robotics have highlighted its potential to handle versatile robotic tasks, current DRL-based collision-free motion planners for manipulators are highly costly, hindering their deployment and application. This is due to an overreliance on the minimum distance between the manipula- tor and obstacles, inadequate exploration and decision-making by DRL, and inefficient data acquisition and utilization. In this article, we propose URPlanner, a universal paradigm for collision-free robotic motion planning based on DRL. URPlanner offers several advantages over existing approaches: it is platform- agnostic, cost-effective in both training and deployment, and applicable to arbitrary manipulators without solving inverse kinematics. To achieve this, we first develop a parameterized task space and a universal obstacle avoidance reward that is independent of minimum distance. Second, we introduce an augmented policy exploration and evaluation algorithm that can be applied to various DRL algorithms to enhance their performance. Third, we propose an expert data diffusion strategy for efficient policy learning, which can produce a large-scale trajectory dataset from only a few expert demonstrations. Finally, the superiority of the proposed methods is comprehensively verified through experiments.