5-Context and AI-based Reasoning for Identification Onboard UxVs (5-CARIBOU)

Context- and AI-Based Reasoning for Identification onBOard UxVs (CARIBOU) is a comprehensive, AI-based automatic target recognition (ATR) agent paired with a fundamentally novel reasoning engine and a state-of-the-art sonar simulator. CARIBOU provides accurate, robust target recognition capabilities that are not easily fooled by novel classes, dim signals, or attempts to deceive, without requiring a human-in-the-loop or burdensome size, weight, and power (SWaP) requirements. In Phase I, we prototyped CARIBOU’s “brain” and demonstrated that it could classify similar contacts with an overall accuracy greater than 97%. In Phase II, we will continue to upgrade Brain while also developing SKULL, which is a novel interface between Brain and the simulator and acoustic processors. Phase II will also consider additional passive sonar sensors and will develop a roadmap for adding other modalities to CARIBOU. Finally, Phase II will target an in-water demonstration using the UD-GARC USV in the Chesapeake Bay. CARIBOU is a joint effort between Boston Fusion Corp, which brings expertise in machine learning research for DoD applications, Advanced Acoustic Concepts, which brings expertise in sonar processing, acoustic modeling, and sonar hardware, and Maritime Applied Physics Corp., which brings expertise in building, modifying, and operating UxVs and other small watercraft.