Comparative performance assessment of path-planning for autonomous mobile robots using Structured Intelligence

Abstract

In this paper the performance of a new algorithm for path-planning of autonomous mobile robots is presented. This algorithm, SIA, is based on the Structured Intelligence Approach, which combines theories on genesis and classification of intelligent behavior in a transdisciplinary manner. The SIA utilizes available knowledge about the environment and handles dynamically changing environments during run-time. Simulations are used to generate the intelligent behavior that solves the path-planning problem. For the simulation environment full implementational details of all SIA modules are discussed. To substantiate the transdisciplinary approach, the performance of SIA was evaluated against the A* algorithm's performance with an empirical procedure, which fulfils social sciences' research schemes: A series of simulation experiments was conducted and data analyses yielded statistically significant effects: SIA shows satisfying results already with no previous knowledge. This means, the SIA is able to generate intelligent behavior. In addition, SIA improves its performance with increasing information. Future research will aim towards real-time implementation and thus complete the simulation's analyses.