HIVE (an HTM-Based Hierarchical Interactive Virtual Environment)

The modern battlefield has become overwhelmingly complex. The sheer volume of data, its spatial, multimodal, and rapidly evolving nature dictate a new approach that allows commanders to more quickly and effectively interpret and interact with data in order to develop a more complete operational picture. Providing tools that enable inspection of space and time correlations to help us classify troop movements at multiple hierarchical levels will become increasingly critical to managing the battlefield in a complex, net-centric environment. We propose to develop the HIVE (Hierarchical Interactive Virtual Environment) prototype, a collaborative C4ISR system that: 1) Uses Hierarchical Temporal Memory (HTM) to automatically recognize patterns of interest in ISR data in order to provide analysis, visualization, and collaboration support to C4ISR analysts. 2) Allows users to easily and flexibly subdivide the Area of Responsibility hierarchically among analysts, both human and automated. Phase I validated HTM’s ability to correctly categorize simple troop movements from GPS and manufactured data. Phase II will apply HTM to realistic Red Force data such as Spot Reports, Observed Position Reports, and other tactical message types.