Topological Navigation of Simulated Robots using Occupancy Grid

TL;DR

This paper introduces a topological navigation method for simulated robots that builds a graph of key locations from an occupancy grid, improving navigation efficiency in a simulated environment.

Contribution

The paper presents a novel topological navigation approach based on occupancy grid skeletonization, enhancing robot navigation efficiency and reducing processing time.

Findings

Significant time savings in navigation process

Effective creation of topological maps from occupancy grids

Improved navigation performance in simulated environments

Abstract

Formerly I presented a metric navigation method in the Webots mobile robot simulator. The navigating Khepera-like robot builds an occupancy grid of the environment and explores the square-shaped room around with a value iteration algorithm. Now I created a topological navigation procedure based on the occupancy grid process. The extension by a skeletonization algorithm results a graph of important places and the connecting routes among them. I also show the significant time profit gained during the process.