Computer-Aided Prognostics and Health Management Using Incipient Failure Detection

One of the main objectives of Industry 4.0 is to build up Smart Factories with improved performance as for productivity, together with lowered maintenance times and costs. In this perspective, Prognostic and Health Management (PHM) is a proactive method to industrial services enhancing maintenance according to the health of the system. PHM entails diagnostic and prognostic engineering tools to recognize the health of the system, and then to choose the prime maintenance actions. The diagnostic tool has to be capable to handle a sizeable volume of data and determine, by means of processing algorithms, the proper set needed for the analysis. The software named MADe can be used as a helpful utility to engineers; it is a model-based toolkit for Reliability, Availability, Maintainability and Safety (RAMS) analysis, capable of optimizing maintenance activities based on the information given by the software, relating to sensor choice and to maintenance strategies. In the PHM framework, the detection of incipient failures is central task of the monitoring the health status of systems that include components sensitive to fatigue or aging. In fact, timely diagnosis allows to schedule maintenance reducing the impact on production outcomes. Based on these considerations, the present paper explains a technique for detecting incipient failures in fatigue sensitive parts, by means of an Equivalent Damage Index (EDI), that can be calculated from the measured signals on the real plant. This procedure is validated, as well as other cutting-edge techniques, to prove its accuracy in detecting incipient breakdowns.