The authors develop an approach to organizing the presentation of large amounts of model-based data in an interactive format patterned after a model of human-human explanatory and argumentative discourse. Portions of this interface were implemented for Honeywell and the Air Force's model-based flight control maintenance and diagnostic system (FCMDS). It is concluded that sufficient information exists in a model-based system to provide a wide range of explanation types, and that, at least in FCMDS, this information is readily accessible though in a format difficult for the user to use. It is also concluded that the discourse approach is a convenient, powerful, and broadly applicable method of representing and controlling information exchange in human-machine, as well as human-human, interactions. >

http://ieeexplore.ieee.org/iel2/653/6713/00270080.pdf

Winning the explanation game; providing user explanations for a model-based expert system

Current flight line maintenance troubleshooting uses static fault isolation trees contained in paper based technical manuals. Future troubleshooting trends will automate the flight line diagnosis of system failures using portable maintenance aides (PMAs). The authors compare and contrast the future automated diagnostic approach to the paper based methodology of today. Field experiment results are also presented that quantify the differences between these two troubleshooting approaches. These field experiments conducted at Luke Air Force Base showed that enhanced maintenance performance was achieved when a diverse work force was equipped with PMAs. >

J.W. Meisner

Papers

Comparison of paper and automated fault isolation techniques