Data Fusion in Support of Spatial Reasoning

Modern situational awareness is supported by a multiplicity of sensors. These provide target information over spatial scales from a few square meters to hundreds of square kilometers. When combined with geographic and other environmental information, the volume of data available to commanders is extensive. This produces challenges in the effective and efficient fusion of all relevant information for maximum utility. Computer architectures offer the throughput to handle large volumes of data. However, automated algorithms are typically ineffective at data fusion, due to their specific and rigid reasoning structures. Hence, the reliance on biologic systems (i.e. human analysts) that employ general and flexible reasoning. Unfortunately, biologic systems are easily overloaded unless data are heavily filtered, decreasing their utility. This program proposes combining general/flexible biologic reasoning with digital computer speed/throughput in automated algorithms. These algorithms fuse spatial target and environment information to perform specific functions, but are not scenario or target specific. Hierarchical systems of fuzzy and crisp rules are employed to perform the processing. Given the data representations and environmental information supported, commercial Geographic Information Systems (GISs) are employed as the framework for such spatial reasoning modules. Current and enhanced GIS functionality to support generic spatial reasoning will be examined. BENEFITS: This research will develop a military spatial reasoning capability which fuses target and environment information, but is not scenario or target specific. Examples include path planning, mobility assessments and unit specialization/generalization. Given the military feasibility of generic spatial reasoning, the concepts can then be applied to commercial uses such as disaster damage assessment, wildlife preservation and law enforcement.