This proposal aims at developing a miniaturized, distributed electroptical sensor system that in real time determines both distance and orientation of a robotic mobile vehicle system with high accuracies relative to a designated area within buildings or factory-type environments. The project will develop a concept originally developed at USC under contract from NASA-JPL. The HexEye system, as it is known, has already been successfully demonstrated to ranges of 10 cm with surfaces with wide ranges of albedo variation. This level of performance is the result of two accomplishments: the measurement based on distribution of reflected light intensity on multiple, symmetrically arranged linear detectors, and data processing based on multiple levels of signal fusion using neural-network type calibration algorithms. The goal of this proposal is to provide a theoretical and experimental evaluation of the feasibility of extending the range of the HexEye by a factor of ten to 1 m, the range needed for mobile robot navigation applications, and to determine the design modifications needed to apply this sensor to mobile robot navigation and docking in tight spaces. The resulting extended sensor system will also be useful for EVA robotics systems and orbital laboratory robotics systems for NASA applications. Autonomous mobile robots are rapidly moving into commercial environments as an attractive alternative to expensive AGV systems based on dedicated, fixed paths. Principal early applications are floor care, security, material transport and research. The commercial partner in this proposal, TRC, has delivered over 300 such autonomous robots to customers around the world. The technology proposed here will lower the cost and improve the performance of these robots, particularly in the ability to navigate in tight spaces with minimal clearance. The first commercial application will be in providing sensing for towed payloads behind an autonomous "tug" in backing into elevators and docking at destination points. This capability opens a potential market for TRC estimated at $500 million.
