First there was reactive maintenance—something broke and it was fixed. Then came preventative maintenance, where things are replaced or repaired before they break based on a manufacturer’s estimate. The next step in organizational efficiency is predictive maintenance.
Thanks to technology advances in data, predictive maintenance tells plant personnel exactly when a machine will need to be fixed. With a predictive maintenance program in place, a company can say goodbye to expensive downtime.
Preventative maintenance vs. Predictive maintenance
Preventative maintenance is developed by manufacturers to help companies avoid unexpected breakdowns. Based on calendar dates or hours of operation, preventative maintenance follows a schedule for repairs. Performing this maintenance can extend the machinery’s operating life.
The trouble with preventative maintenance is it a one-size-fits-all solution. Preventative maintenance is better than waiting for machinery to break before repairing it, but it still leaves a lot to be desired. Based on theoretical rates of failure, preventative maintenance doesn’t account for the actual performance of the machine.
Predictive maintenance, on the other hand, monitors equipment in real-time. Instead of scheduling maintenance on a machine after 200 hours of service just because, it will analyze the machine’s performance and warn the maintenance team when it will need service, perhaps weeks in advance. This analysis would be based on sensors measuring things like lubrication, vibration and corrosion.
With predictive maintenance, a highly utilized line motor will be repaired based on its usage. Lower-utilized equipment would likewise receive less attention.
The ingredients of a predictive maintenance program
Implementing a predictive maintenance program basically requires three things: data, time and analysis.
Start by connecting equipment. Then, with the help of sensors, the system can constantly collect data on the status of the plant’s equipment. Internet of Things (IoT) connectivity brings the data from the machines into a central computer system. Once you have a database with streams of data coming in, the stage is set for machine learning—where the computer system analyzes the data, learns from it and makes its own recommendations.
There may be quite a bit of time between starting to gather data and the computer learning enough to make recommendations. It takes time to collect enough data, and while you’re collecting the initial data, continue the usual preventative maintenance.
The powerful thing about machine learning and big data is its iterative nature. Every bit of data the system gets leads to a slightly more refined model and better prediction capabilities.
Benefits of predictive maintenance
Companies that implement predictive maintenance reap harvests of less downtime, better decision-making, more efficient maintenance and lower total cost of ownership (TCO). When implemented, a predictive maintenance program can alert plant managers and procurement staff when machines will need service, months or weeks in advance so you can always be sure to meet manufacturing deadlines.
With predictive maintenance, companies benefit from better resource exploitation. Instead of replacing items with plenty of operating life left because that’s what the schedule says to do, predictive maintenance allows companies to get the maximum amount of use from a machine and then schedule a cost-effective time to conduct preemptive repairs.
Downtime is outrageously expensive—especially when it’s unplanned. There’s the lost productivity, emergency maintenance, expedited shipments of parts, perhaps spoiled product as well as the cost of maintenance and procurement staff addressing the crisis instead of whatever they were scheduled to do that day.
Predictive maintenance requires advanced technology to manage the sensor fleet deployed on the equipment. This mirrors the building automation capabilities taking the building and operations world by storm. The result is a tighter integration between maintenance and operations, which can affect more efficient labor.
Maintenance is always cheaper than downtime and with predictive maintenance, resources like maintenance staff can be allocated far more efficiently.
IM Supply is a national electric and lighting MRO partner with deep experience in helping customers leverage lighting for maximum plant productivity and lowering total cost of ownership. IM Supply can offer a wealth of advice on keeping plants running safely, more efficiently and in full compliance with applicable regulations.