Soldier-Centric Knowledge Engineering: Lessons from TED

Developers of expert system technologies often struggle with the problem of how best to couple the art of knowledge acquisition and knowledge engineering into a useable system. When combined properly, the results obtained are outstanding (as evidenced by the numerous successful systems produced in medicine, finance, and other disciplines). On the other hand, when uniting these processes does not occur properly, no amount of money or time can stay the ultimate outcome-a system whose results fall short of the expected benefits and trigger the dreaded trying-and-failing syndrome. For a team of scientists from the US Army Research Laboratory (ARL) and the US Army Ordnance Center and School (OC&S), trying and failing was never an option. The Ordnance School's Directorate of Combat Development assigned the team the task of developing an expert diagnostic system to troubleshoot and repair the Army's Abrams main battle tank. The system, known as TED (turbine engine diagnostics), gives Army tank mechanics the ability to effectively and efficiently diagnose faults, perform necessary repairs, order parts, validate serviceability, and maintain necessary maintenance records. It also provides a comprehensive online tutorial suite. The TED program became the first successful Army program to combine technologies from artificial intelligence with maintenance doctrine into a fielded Army system