Prediction of the remaining useful life of mechatronic systems, using internal sensors

In this work, a Condition Monitor (CM) procedure has been applied to an automatic machine for cutting of steel bars, for the purpose of estimating the Remaining Useful Life (RUL) of specific components of the system. Two sensors inside the system and an external one have been used to monitor the wear condition of the blade in the cutting unit. Experimental data have been processed to extract synthetic features and, on the basis of those, a fitting Artificial Neural Network (ANN) has been trained and tested. The preliminary results appear to be interesting, showing a satisfactory ability of the ANN to identify the number of working cycles. This study represents a first step towards the ultimate goal of improving the maintenance strategies of the automatic machine.