Hybrid Control of a Segway Platform Developed in MRDS

Abstract

<p>A Segway RMP200 has been bought by Victoria University for the purpose of making an autonomous robot. The focus of this project was to create reusable services that use existing navigation algorithms to control the Segway within an indoor environment. A SICK LMS100 laser rangefinder was added to detect obstacles and allow localization of the Segway within a known map. A hybrid navigation algorithm consisting of an A* path planner with a dynamic window is used for motion planning and obstacle avoidance. The control system followed a Service Oriented Architecture implemented in Microsoft Robotics Studio using the C# .NET programming language. Four services were created during the project to interface with the SICK LMS100 scanner, control the Segway RMP200, implement the hybrid navigation algorithm and provide a graphic user interface for the system. Tests show that the Segway is able to navigate and maintain localisation within the operating environment by identifying and associating corner and door landmarks within the environment.</p>