Trusted AI for Multimodal MTI

Moving target engagement is a Department of Air Force’s (DAF) operational imperative. Combining multimodal space sensor data (e.g., EO/IR and SAR) will produce higher moving target indication (MTI) accuracy robust to operational uncertainties. Reducing the cognitive load on intelligence analysts while establishing AI trust is another key performance metric for DAF. This Phase I proposal addresses these problems with new AI model design, analysis, and software demonstration. The proposed R&D approach will deliver 10X higher MTI accuracy and significant human-AI trust. The delivered software product will overcome the limitations of current solutions that are manual, error-prone, and suffer from high latency. Several DAF stakeholders and commercial partners have indicated strong interest in the proposed software solution as evidenced in the Letters of Support.