Implementing Deep Learning Model to Predict the Maintenance of an Elevator System

In this world, where everything is becoming automated, humans have also got used to this fast-paced life. One of the machines that we widely use in our day-to-day lives are elevators. Elevator systems have become an important part of the buildings. Hence, in order to perform well, even elevators require proper maintenance. Therefore, it becomes very much essential to strengthen the maintenance procedure of elevators. Here, the role of predictive maintenance comes into play. It helps to find out the proper time period, in which the maintenance should be done without wasting the maintenance cost along with Remaining Useful Life of elevator. The technique used here is deep learning. It gives us finer outcomes because of the new deep attributes, that are extracted from the dataset in contrast with already existing attributes. This work consists of implementing a deep learning algorithm for fault detection and prediction of an elevator system. LSTM has been implemented on the dataset, which consisted of sensor data which was recorded on the basis of maintenance actions being taken. The experimental results present mean absolute error of about 7.05%.