Anytime Planning: A Motion Planner for Dynamic Environment

TL;DR

This paper introduces a novel probabilistically complete motion planning algorithm for dynamic environments, optimizing for minimal computation time or path cost within given constraints, suitable for autonomous vehicles and multi-robot systems.

Contribution

It presents a new sequential BIT* algorithm that outperforms existing methods like DOVS in dynamic obstacle-rich environments, considering dynamic constraints for feasible paths.

Findings

Outperforms DOVS in path length and computation time

Plans feasible paths considering dynamic constraints

Proven effective through extensive simulations

Abstract

Motion planning in the presence of multiple dynamic obstacles is an important research problem from the perspective of autonomous vehicles as well as space-constrained multi-robot work environment. In this paper, we address the motion planning problem for multiple dynamic obstacle rich environment and propose a probabilistically, complete novel motion planning algorithm. Our claim is that given a fixed path cost i.e. the Euclidean path length, the proposed algorithm plans a path with the least computational time as compared to the state-of-the-art techniques. At the same time, given the time duration for planning, it plans the minimum cost path. Dynamic constraints have been taken into consideration while designing the planner such that the optimal planned path is feasible for implementation. The results of extensive simulation experiments show that the proposed sequential BIT* outperforms the DOVS both in the planned path length as well as the path generation time.