Network Interfaces and Fuzzy-logic Control for a Mecanum-wheeled Omni-directional Robot

This paper presents the results of a project to develop an omni-directional robotic system at NASA Langley’s Robotics and Intelligent Machines Lab. The robot, named Mobius, was created as part of an investigation into omni-directional motion (using four Mecanum wheels) and fuzzy-logic control. The robot uses a highlevel processor running Linux to connect to wireless networks, process user-input data, relay camera imagery, and communicate with a low-level microcontroller. All the programs for user input and imagery viewing are cross-platform, Internet-enabled, client-server applications. When within range of a wide-area-network wireless access point, Mobius can be driven from an arbitrary location on the Internet. Low-level control of Mobius’ motors is through a simple microcontroller-based fuzzy-logic algorithm. The robot platform is capable of translation in any direction, bidirectional rotation, and simultaneous translation with rotation. Future work includes constructing a vehicle with three Mecanum wheels and autonomous algorithms. The algorithms will fuse sonar ranging with stereoscopic and panoramic imagery for autonomous robot navigation and topologic robot localization.