Signal understanding and tool condition monitoring

This paper presents a new method, called signal understanding, for the monitoring of engineering processes, in particularly for the monitoring of tool conditions in machining processes. The new method is based on the blackboard system, an artificial intelligence method developed in the 1980s. It emulates a group of experts examining sensor signals from various aspects, and making monitoring decisions step-by-step. The blackboard system consists of two blackboards: an event blackboard, used to track the interpretations of the signal by the experts and a control blackboard, used to direct the interpretation process leading to monitoring decisions. As demonstrated by an example of tool condition monitoring in end milling, the method has a number of advantages over the existing methods, including improved reliability and reduced decision time. (C) 1999 Elsevier Science Ltd. All rights reserved.