Ant Colony Based Algorithm to Identify Alarm Patterns For Predicting Equipment Failures

The core of condition based maintenance is the ability to analyze data in real-time and to predict potential equipment failures. The goal is to avoid catastrophic failures through prompt execution of remedial actions. This paper describes an attempt to develop an Ant colony based algorithm, called Ant Predictor, for identifying alarm pattern for prediction of equipment failures. We collected event logs containing various fault alarms and system activity reports from an automatic material handling equipment for a period of two years. By mining the history data with the Ant Predictor, it is possible to find a plausible sequence of alarms for predicting equipment failure. The initial results demonstrate the feasibility of the proposed Ant Predictor for prediction of equipment failures.