A Connectivity-Based Method for Enhancing Sampling in Probabilistic Roadmap Planners

The motion planning is a difficult problem but nevertheless, a crucial part of robotics. The probabilistic roadmap planners have shown to be an efficient way to solve these planning problems. In this paper, we present a new algorithm that is based on the principles of the probabilistic roadmap planners. Our algorithm enhances the sampling by intelligently detecting which areas of the configuration space are easy and which parts are not. The algorithm then biases the sampling only to the difficult areas that may contain narrow passages. Our algorithm works by dividing the configuration space into regions at the beginning and then sampling configurations inside each region. Based on the connectivity of the roadmap inside each region, our algorithm aims to detect whether the region is easy or difficult. We tested our algorithm with three different simulated environments and compared it with two other planners. Our experiments showed that with our method it is possible to achieve significantly better results than with other tested planners. Our algorithm was also able to reduce the size of roadmaps.