Message anticipation applied to collaborating unmanned underwater vehicles

Researchers at the University of Idaho (UI) have developed a fleet of unmanned underwater vehicles (UUVs) that autonomously and collaboratively engage in mine-countermeasure (MCM) and magnetic signature assessment (MSA) missions. The use of intervehicle acoustic messages is instrumental in successful completion of these types of missions. In what follows, we argue that enabling the UUVs to anticipate these acoustic messages would improve their ability to successfully complete missions. We begin by describing the UUVs, their acoustic communications language, known as AUVish-BBM, and associated protocol, and the specifications of MCM and MSA missions. We continue by reviewing published literature on the use of anticipation by humans in natural language. Paying careful attention to both the structure of natural language and the way humans anticipate passages of it, we outline an approach to UUV message anticipation in AUVish-BBM, based on a UI-developed paradigm called Language-Centered Intelligence (LCI). In the LCI message anticipation paradigm, existing language and behavioral logics are deployed hypothetically to anticipate messages forthcoming from other vehicles in the fleet. We conclude by describing and detailing results from error-correction analysis that provides the basis for a logic governing the Message Anticipation Module, currently under development at UI.