Automatic Path Planning of Industrial Robots Comparing Sampling-Based and Computational Intelligence Methods

Abstract

In times of industry 4.0 a production facility should be smart. One result of that property could be that it is easier to reconfigure plants for different products which is, in times of a high rate of variant diversity, a very important point. Nowadays in typical robot based plants, a huge part of time from the commissioning process is needed for the programming of collision free paths. This mainly includes the teach-in or offline programming (OLP) and the optimization of the paths. To speed up this process significantly, an automatic and intelligent planning system is necessary. In this work we present a system which can plan paths for industrial robots. We compare widely used sampling-based methods like PRM or RRT with Computational Intelligence (CI) based methods like genetic algorithms.