Anytime Motion Planning using the RRT*

Abstract

The Rapidly-exploring Random Tree (RRT) algorithm, based on incremental sampling, efficiently computes motion plans. Although the RRT algorithm quickly produces candidate feasible solutions, it tends to converge to a solution that is far from optimal. Practical applications favor "anytime" algorithms that quickly identify an initial feasible plan, then, given more computation time available during plan execution, improve the plan toward an optimal solution. This paper describes an anytime algorithm based on the RRT * which (like the RRT) finds an initial feasible solution quickly, but (unlike the RRT) almost surely converges to an optimal solution. We present two key extensions to the RRT * , committed trajectories and branch-and-bound tree adaptation, that together enable the algorithm to make more efficient use of computation time online, resulting in an anytime algorithm for real-time implementation. We evaluate the method using a series of Monte Carlo runs in a high-fidelity simulation environment, and compare the operation of the RRT and RRT * methods. We also demonstrate experimental results for an outdoor wheeled robotic vehicle.

Efficient Sampling-Based Approaches to Optimal Path Planning in Complex Cost Spaces

“…The conditional activation heuristic consists of planning with a regular RRT until the first solution is found, and only then activating the procedures specific to RRT* [11]. The branch-and-bound heuristic consists of trimming the nodes in G that cannot allow finding paths with costs lower than that of the current solution path, which is assessed using a cost-to-go function [13]. Both heuristics are beneficial on the Transport and Snake problems.…”

“…Another variant of RRT, called RRT*, has been devised to solve the optimal path planning problem [12]. RRT* has been shown to guarantee asymptotic optimality, and has been applied to various robotic problems [11][12][13]. However, it might converge slowly in high-dimensional spaces [4].…”

“…Historically, the first cost criterion to be considered was path length [1], [6], [12,13], [17,18]. More interesting problems can be addressed with more sophisticated criteria, based on the definition of a cost function over the configuration space, which then becomes a cost space.…”

Towards a Trajectory Planning Concept: Augmenting Path Planning Methods by Considering Speed Limit Constraints

“…A combination of both (RRTs and Probabilistic Roadmaps) can be found in [32]. The RRT* algorithm [16] overcomes the issue with less optimal solutions, which is the main shortcoming of RRTs. RRT* follows the asymptotic optimality property, which guarantees convergence to optimal solutions (paths).…”

Sampling-Based Motion Planning Algorithms for Replanning and Spatial Load Balancing

“…At the core of this line of work are the PRM* and RRT* [29]. The RRT* handles any-time applications [30] and manipulators [50]. The RRT* paths can be composed of many more nodes than is strictly necessary.…”