Knowledge Acquisition Tools for Cognitive Design Aid and Development

Rotorcraft Pilot's Associate (RPA) flight test and other previous and ongoing system evaluations [Dorn99] have demonstrated the military utility of Cognitive Decision Aiding Systems (CDASs). There is, however, an increasing awareness that the knowledgewhich must be acquired, captured, represented, implemented and tuned offers a tremendous cost and process hurdle to the development, evolution and fielding of CDA Systems. What is needed is a means of tracing the lifecycle of every bit of knowledge in thesystem, and of every design decision that involves that knowledge-and of doing so in a manner that does not impose undue additional workload on the engineers building the system. In this effort we will examine the benefits realizable from implementing anumber of alternative proposed extensions to the Unified Modeling Language (UML) use case formalism. We anticipate melding attributes from a number of available extension proposals into a synthesized methodology enhancement that effectively integratesacquired knowledge in support of CDASs into the standard software development process. We will apply this methodology to a representative set of problem cases experienced on RPA and other similar efforts, analyze the contributions and refine the design.Other supporting advances in knowledge acquisition and tagging are described.Reduced cost of knowledge acquisition activities. More effective use of acquired knowledge. Better insight into influence effects of knowledge on standard software developmentenvironments. Capability to build critics analogous to existing software inspection tools through enhanced traceability.