Sampling-Based Path Planning for Tethered Robot Chains

Abstract

Motivated by human-chains in rescue missions, this paper proposes a scalable path planning algorithm for multiple mobile robots that are tethered to one another in a chain topology with finite-length tethers. Specifically, our approach trades off optimality for scalability and computational tractability by adding some simplifying, yet realistic constraints that can significantly reduce computation. In particular, by maintaining the existence of tether configurations that coincide with collision-free, feasible paths for the robots, we remove the need to check that the tether configurations are collision-free, which is often a bottleneck since the tethers are infinite-dimensional. Our proposed path planning framework for tethered robot chains builds upon sampling-based algorithms such as RRT*, BIT*, and ABIT*. Finally, we prove the probabilistic completeness of the approach, ensuring reliable path generation, and demonstrate the effectiveness of our approach in simulation experiments.