An advanced unmanned vehicle for remote applications

摘要

An autonomous mobile robotic capability is critical to developing remote work applications for hazardous environments. A few potential applications include humanitarian demining and ordnance neutralization, extraterrestrial science exploration, and hazardous waste cleanup. The ability of the remote platform to sense and maneuver within its environment is a basic technology requirement which is currently lacking. This enabling technology will open the door for force multiplication and cost effective solutions to remote operations. The ultimate goal of this work is to develop a mobile robotic platform that can identify and avoid local obstacles as it traverses from its current location to a specified destination. This goal directed autonomous navigation scheme uses the Global Positioning System (GPS) to identify the robot`s current coordinates in space and neural network processing of LADAR range images for local obstacle detection and avoidance. The initial year funding provided by this LDRD project has developed a small exterior mobile robotic development platform and a fieldable version of Sandia`s Scannerless Range Imager (SRI) system. The robotic testbed platform is based on the Surveillance And Reconnaissance ground Equipment (SARGE) robotic vehicle design recently developed for the US DoD. Contingent upon follow-on funding, future enhancements will develop neural network processing of the range map data to traverse unstructured exterior terrain while avoiding obstacles. The SRI will provide real-time range images to a neural network for autonomous guidance. Neural network processing of the range map data will allow real-time operation on a Pentium based embedded processor board.