Cognitive Architecture “Thinks Through” Changing Conditions, Offers Solutions Published March 10, 2014 By Holly Jordan, AFRL/RQOB Aerospace Systems WRIGHT-PATTERSON AIR FORCE BASE, Ohio -- The Air Force Research Laboratory has demonstrated a new computational intelligence concept that could enable adaptation to varying aircraft conditions and real-time cognitive assessment about the state of a vehicle.This capability could open the door to more autonomous and efficient management of vehicle operations. Conscious Architecture for State Exploitation (CASE) is a general-purpose, computational decision-making framework that generates information about the state of a system and makes adaptive decisions based on a variety of factors. It is different from other decision-making architectures in that it is designed to mimic aspects of human cognition, essentially functioning as the "brain" of the aircraft. Traditional computational intelligence frameworks are somewhat limited as to the types and scope of problems they can effectively handle. CASE, on the other hand, functions more like true cognition. In essence, it "thinks through" the conditions at hand, adapting as those conditions change and choosing the most appropriate solution. AFRL tested CASE in a laboratory setting to monitor the integrity of a representative airframe component under flight-like fatigue loading conditions. As the load cycles accumulated and the component began to fracture, CASE automatically generated a timely maintenance recommendation. Over tests simulating 1000 component lifetimes, a 900 percent time increase between maintenance actions and a 10.7 percent increase in availability were observed over the traditional practice of predetermined manual inspections. Additionally, CASE was shown to decrease both component ownership cost and component downtime by 79 percent. Such dramatic efficiency improvements could revolutionize the future of vehicle maintenance and management. Following this successful test, researchers will extend the study of CASE for other Air Force autonomy applications, such as unmanned system autonomous terminal area and ground operations.