|
useful cases from previous editions General Motors Tries to Find Machine Failures Before They Occur Charley is a computerized diagnostic system that helps mechanics decide whether a catastrophic machine failure will occur soon. It also helps them identify the likely cause. After a repair, mechanics use Charley to diagnose whether the repair solved the original problem. Charley was designed to imitate the thought processes of Charles Amble, a maintenance engineer who had become an acknowledged industry expert during his 23 years with General Motors. The system has been installed at many plants and has been used successfully to reduce machine repair costs. Charley uses three types of information. The first is more than 1,000 rules describing the way Charles Amble would perform a machine diagnosis. The second is a description of the major components of a machine and the relationship between these components. The third is a set of current vibration signatures consisting of frequencies and amplitudes of vibration patterns measured at different points on the machine's surface. These measurements are taken with a handheld accelerometer and recorder. Charley uses the rules and machine-description information to analyze the vibration information. Mechanics use Charley in interactive sessions in which the mechanic describes the symptoms and Charley asks questions required to develop and validate hypotheses about the nature of the problem. Questions:
Source: Bajpai, Atul, and Richard Marczewski. Charley: An Expert System for Diagnostics of Manufacturing Equipment. In Herbert Schorr and Alain Rappaport, Proceedings of the First Annual Conference on Innovative Applications of Artificial Intelligence. Menlo Park, Calif.: American Association for Artificial Intelligence, 1989, pp. 178-185.
Back to Useful Cases from Previous Editions
|