Absolute Localization in GPS-denied Environment for Autonomous Unmanned Ground and Micro-air Vehicle Systems

Polaris proposes to make use of currently available sensor suites on AUGV/MAV platforms to implement a Kalman filter-based integrated solution to the autonomous localization problem in GPS-denied environments, titled Stereo Camera Absolute Localization and Map correlation System (SCALMS). This effort will focus on the development of two distinct subsystems, dead reckoning and absolute positioning. The innovation lies in the implementation of the dead reckoning localization subsystem and also in the novel methods for correlating maps in the absolute positioning subsystem. The proposed Phase I will result in the implementation and field test of the first subsystem and the analysis and design of the second. A design which integrates the two into the SCALMS architecture is provided. Finally, novel methods for handling feature-poor and feature-rich environments, as well as localization reset, are suggested. We estimate that the SCALMS architecture could be implemented on many AUGV platforms in production today without adding additional sensing hardware and without exceeding SWAP, processing, and memory requirements. The final system will be ideal because of its integration of some proven technologies, but innovative due to its application of others.