A REVIEW ON MAINTENANCE TECHNIQUES FOR INDUSTRIAL EQUIPMENT AND ITS MACHINE LEARNING ALGORITHMS

Maintenance is a complex issue for companies with lot of equipment. For maintenance companies depend on scheduled maintenance to repair the fault and damaged parts. In manufacturing industries, machines are broadly classified into rotary and reciprocating. The integral part of rotary machine is electric motor. The continuous evaluation of the health of the motor has to be assessed to increase the lifetime of the equipment. So, it is very much necessary to prevent sudden failure of the equipment. With the emerging technology of industrial internet of things, machine learning and transmission protocols such as MQTT (Message Queuing Telemetry Transport) paved a way for predictive maintenance by monitoring the real-time operation of the machine that avoids unexpected breakdown, preventing unpredictable losses and decreases the maintenance costs which extends the equipment life. A review on recent advancements for collecting the industrial motor inputs, outputs, and load and parts data such as temperature, vibration, and acoustics using various industrial sensors for analysing real-time state of the machine are highlighted in this paper. To build an algorithmic model for predicting the maintenance data of an industrial machine using machine learning technique and highlighting the research works, which provides the basement for carrying out future research work.