A Navigation and Decision Making Architecture for Unmanned Ground Vehicles: Implementation and Results with the Raptor UGV

Abstract : Researchers at Defence R&D Canada- Suffield, under the Autonomous Land Systems (ALS) and Cohort projects, have been working to extend/enhance the capabilities of Unmanned Ground Vehicles (UGVs) beyond tele-operation. The goal is to create robotic platforms that are effective with minimal human supervision in outdoor environments. This report is a summary of the progress made in high level vehicle control, specifically the implementation and testing of algorithms providing point-to-point navigation and decision making capabilities for UGVs. To reach goals by traversing unknown terrain requires a number of navigation functions, including path tracking, obstacle avoidance, path planning and decision making modules. This report presents details of the theoretical underpinnings, the software design architecture, and results of implementing autonomous navigation and decision making software on a robotic platform, given competing priorities and limited sensing technologies.