Motion planning in high-dimensional spaces

TL;DR

This paper reviews various strategies for high-dimensional motion planning in robotics, comparing their strengths and weaknesses, and discusses future research directions in real-time, complex environment navigation.

Contribution

It provides a comprehensive overview and comparison of grid search, sampling, and trajectory optimization methods for high-dimensional motion planning.

Findings

Sampling-based methods are scalable to high dimensions.

Trajectory optimization offers smooth and feasible paths.

Grid search methods are less practical in high-dimensional spaces.

Abstract

Motion planning is a key tool that allows robots to navigate through an environment without collisions. The problem of robot motion planning has been studied in great detail over the last several decades, with researchers initially focusing on systems such as planar mobile robots and low degree-of-freedom (DOF) robotic arms. The increased use of high DOF robots that must perform tasks in real time in complex dynamic environments spurs the need for fast motion planning algorithms. In this overview, we discuss several types of strategies for motion planning in high dimensional spaces and dissect some of them, namely grid search based, sampling based and trajectory optimization based approaches. We compare them and outline their advantages and disadvantages, and finally, provide an insight into future research opportunities.