ESPRIT: Schemas for Intent Recognition
Department of Defense
Agency Tracking Number:
Solicitation Topic Code:
Small Business Information
Soar Technology, Inc.
3600 Green Court, Suite 600, Ann Arbor, MI, -
Socially and Economically Disadvantaged:
CEO and Chief Scientist
CEO and Chief Scientist
AbstractThe potential for force multiplication in Unmanned Aircraft Systems (UASs) rises with the degree to which they can be"neglected"by human controllers and operators. There is currently a great deal of human cognitive attention required to manage UAS assets safely and effectively, particularly in cluttered environments such as the terminal area of operations (TAO). To improve UAS autonomy, SoarTech, with support from the University of Michigan (UM), will design and implement a system for Explanation, Schemas, and Prediction for Recognition of Intent in the TAO (ESPRIT). ESPRIT provides an intelligent situational understanding capability for UASs that uses knowledge and reasoning techniques to infer the intent of other aircraft. At the core of ESPRIT"s executable system will be the Soar cognitive architecture, providing efficient implementation of memory management, pattern matching, and goal-based reasoning. ESPRIT will incorporate cognitive capabilities for situation understanding, exception and conflict handling, and task interleaving, as well capabilities for plan monitoring and plan recognition using a Soar-based"Schema Engine". ESPRIT will also incorporate a cognitive explanation mechanism to generate potential explanations of track behavior in those anomalous situations where the system"s other mechanisms provide only partial explanations of observed behaviors. BENEFIT: ESPRIT provides an intelligent situational understanding capability for UASs that uses knowledge and reasoning techniques to infer the intent and future actions of aircraft in the TAO. This will lead to cheaper, safer, and more autonomous future UAS systems. In addition, the capability will enhance future UAS operator and ATC user interfaces, improving overall mission effectiveness and reducing errors and conflicts. There are three primary classes of application that the current research and development will benefit: on-board decision-support systems (for direct control of UASs and assistance for piloted aircraft), decision-support and improved visualization for UAS operators, and decision-support and improved visualization for ATCs. In addition, the ESPRIT capabilities will enhance the capabilities of computer-generated forces for DOD simulations for experimentation, acquisition, and training, improving the quality and ratio of CGF populations for human operators.
* information listed above is at the time of submission.