Autonomous Mobile Surveillance Based on RTK-GPS in Urban Canyons

This paper describes an autonomous mobile surveillance used in plants in high-rise buildings. This consists of a wireless LAN, a base station, and an autonomous vehicle. The vehicle uses GPS/INS navigation using network-based Real-Time Kinematic GPS (RTK-GPS) with Positioning Augmentation Services (PASTM, Mitsubishi Electric Corporation 2003), an Area Laser Radar (ALR), a slave camera, and an Omni-Directional Vision (ODV) sensor for surveillance and reconnaissance. The vehicle switches control modes - normal, road tracking, and crossing recognition - based on vehicle navigation error. A field test shows that the vehicle tracks a planned straight paths within 0.10m accuracy and planned curved paths within 0.25m even without RTK fixed solutions. Field experiments and analysis prove that the proposed navigation provides sufficient navigation and guidance accuracy under poor satellite geometry and visibility, and that the panorama image database with absolute positioning is useful for surveillance.