Experience-based Subproblem Planning for Multi-Robot Motion Planning

TL;DR

This work proposes a novel approach that averages experience-based planning by constructs and utilizes databases of solutions for smaller sub-problems, allowing for efficient handling of more complex MRMP scenarios.

Abstract

Multi-robot systems enhance efficiency and productivity across various applications, from manufacturing to surveillance. While single-robot motion planning has improved by using databases of prior solutions, extending this approach to multi-robot motion planning (MRMP) presents challenges due to the increased complexity and diversity of tasks and configurations. Recent discrete methods have attempted to address this by focusing on relevant lower-dimensional subproblems, but they are inadequate for complex scenarios like those involving manipulator robots. To overcome this, we propose a novel approach that %leverages experience-based planning by constructs and utilizes databases of solutions for smaller sub-problems. By focusing on interactions between fewer robots, our method reduces the need for exhaustive database growth, allowing for efficient handling of more complex MRMP scenarios. We validate our approach with experiments involving both mobile and manipulator robots, demonstrating significant improvements over existing methods in scalability and planning efficiency. Our contributions include a rapidly constructed database for low-dimensional MRMP problems, a framework for applying these solutions to larger problems, and experimental validation with up to 32 mobile and 16 manipulator robots.

Figures (8)

• Figure 1: A simplified overview of our method: a) Detect a conflict between two robots' paths. b) Define a local subproblem around the conflict. c) Retrieve the best solution from the database and solve the subproblem.
• Figure 2: Construction of the subproblem solution database: (a) For mobile robots, a reduced boundary is used to generate random subproblems involving 2, 3, and 4 robots. (b) For fixed-base manipulators, random subproblems for 2 robots are created for various arrangements, including horizontal, vertical, and diagonal
• Figure 3: Multi-mobile robot scenarios: Robots move from random start positions to random goal positions in environments both without (a) and with (b) random obstacles.
• Figure 4: Multi-manipulator scenarios: Robots move from random start positions to random goal positions in environments both with (b) and without (a) varying obstacles.
• Figure 5: Results for the scenario of mobile robots without obstacles