RRT-CoLearn: towards kinodynamic planning without numerical trajectory optimization

TL;DR

This paper introduces RRT-CoLearn, a kinodynamic planning method that uses indirect optimal control for data generation, enabling learning of both distance metrics and steering inputs, significantly speeding up planning without local planners.

Contribution

It proposes using indirect optimal control for data generation in learning RRTs, reducing computational effort and eliminating the need for a local planner.

Findings

10-fold speed-up in data generation and planning time

Eliminates the need for a local planner in learning RRTs

Achieves efficient kinodynamic planning on a pendulum swing-up task

Abstract

Sampling-based kinodynamic planners, such as Rapidly-exploring Random Trees (RRTs), pose two fundamental challenges: computing a reliable (pseudo-)metric for the distance between two randomly sampled nodes, and computing a steering input to connect the nodes. The core of these challenges is a Two Point Boundary Value Problem, which is known to be NP-hard. Recently, the distance metric has been approximated using supervised learning, reducing computation time drastically. The previous work on such learning RRTs use direct optimal control to generate the data for supervised learning. This paper proposes to use indirect optimal control instead, because it provides two benefits: it reduces the computational effort to generate the data, and it provides a low dimensional parametrization of the action space. The latter allows us to learn both the distance metric and the steering input to connect two nodes. This eliminates the need for a local planner in learning RRTs. Experimental results on a pendulum swing up show 10-fold speed-up in both the offline data generation and the online planning time, leading to at least a 10-fold speed-up in the overall planning time.