ACCURATE TO A FAULT: A NEW, SOFTWARE‐ORIENTED APPROACH TO MACHINE HEALTH MONITORING

Predicting and circumventing incipient faults in expensive and intensively used machinery becomes more important every day — from the points of view of extending machinery life, of reducing costly downtime and of avoiding critical failure situations. The author describes recent work done by one of Britain's leading scientific engineering consultancies in collaboration with the Naval Engineering Department of The Royal Aerospace Establishment. A series of techniques was evolved which, when integrated into an intelligent system, can produce an effective and economic condition‐monitoring concept for supervising turbocharged diesel engines. Based on the processing of signals from conventional sensors it could be added to a machine control system without significantly increasing its complexity.