Cognitive Architecture Bridge (CAB)

Years of research in human behavior modeling have yielded a wide array of cognitive architectures. The most vital challenges confronting the field are to find an economical way to leverage the strengths of each of the multiple existing architectures and to facilitate and simplify the manner in which new cognitive models are produced, as well as to reduce the time entailed in their production. Requirements such as cultural effects add another layer of complexity that needs to be addressed. The Air Force is investigating new integrative architectures that incorporate more realistic human and organizational behavior. CHI Systems proposes to address this challenge by developing a new meta-architecture intended to serve as a bridge between existing cognitive modeling architectures. Specifically, CHI proposes the "Cognitive Architecture Bridge" (CAB), using a blackboard-system approach to combine different pre-existing components working at different levels of abstraction. CAB will give model developers the ability to assemble existing components from supported cognitive architectures, together with the ability to represent multiple levels of abstraction and cultural effects. Phase I products will include a design for the CAB meta-architecture, a proof-of-concept and demonstration model interface with CAB, an assessment of non-conventional architectures, and a Phase II plan.