NASA 1996 SBIR Phase I


PROPOSAL NUMBER : 96-1 01.14-3474

PROJECT TITLE : Agent-Based Performance Assessment Tool for General Aviation Operations Under Free Flight

TECHNICAL ABSTRACT (LIMIT 200 WORDS)

Although considerable effort is being devoted to understanding the effect of free flight air traffic management (ATM) concepts on commercial aviation, the consequences for general aviation (GA) have received relatively less attention. We thus propose to develop a computational tool to assess free flightÕs potential GA impact, combining an agent-based representation of the overall pilot/vehicle/ATM system with model-based metrics of pilot situation awareness (SA) and performance. We will focus on the single GA pilot, and demonstrate overall feasibility of generating objective SA and error metrics for evaluating procedures design options, across a range of free flight scenarios. The approach relies on a multi-stage agent-based representation of the pilot's SA and decision-making behavior, and a multidimensional metric that reflects SA, performance, and error propensity. Bayesian networks support SA computations, and an object-oriented implementation facilitates integration with other software packages; a graphical user interface supports scenario specification and on-line performance analysis. We propose to develop a prototype tool, apply it to the free flight environment, integrate it with a realistic ATM model, and demonstrate its use in design option evaluation.
POTENTIAL COMMERCIAL APPLICATIONS
The proposed technology will directly support closing the human factors loop on the development of free flight ATM concepts. The underlying modeling approach will support human factors assessments in other domains (e.g., nuclear power plants, chemical processing centers, power dispatch stations, intelligent vehicle highway system control centers, etc.) where it is desirable to determine how system design affects operator awareness and performance before committing to a specific configuration. We plan to extract the belief network algorithm that emulates pilot situation assessment in the presence of uncertainty, and develop a commercial-quality software tool for computer-based reasoning under uncertainty.
NAME AND ADDRESS OF PRINCIPAL INVESTIGATOR
GregL. Zacharias
55 Wheeler St.
Cambridge, MA 02138
NAME AND ADDRESS OF OFFEROR