Vibration analysis and predictive maintenance on gearbox of lathe machine using machine learning

In the Industry 4.0 era, the use of digital technologies to improve manufacturing chain efficiency and product quality through connectivity and digitization. The goal of this study is to use predictive maintenance using machine learning methods to avert unexpected engine. Maintenance strives to reduce and eliminate the number of failures that occur during manufacturing. Predictive Maintenance is the practice of monitoring the status of rotating equipment in order to discover early failures and prevent catastrophic breakdowns in critical components of a manufacturing chain. The implementation of predictive maintenance in machine learning algorithms can be used to perform lifetime detection activities with rotation (rpm) and usage time parameters. Machine learning can make predictions based on a specified data set. The results showed that the remaining service life of the gearbox component was 8411 hours and the vibration conditions produced by the gearbox were still at a satisfactory level according to the ISO 10816. Linear regression algorithms established their capacity to forecast a value by inputting RPM data and the length of time of use. We calculate the gearbox’s RUL is 8414 hours. format