Interest in predictive maintenance continues to grow, for a variety of reasons: the Industrial Internet of Things (IIOT), increasing use of sensors, higher volumes of data, and easier processing and analytics, to name a few. Unfortunately, that means many people expect everyone to already know everything about predictive maintenance.
What if you’d like to understand it from the ground up, with definitions and examples? If so, this post is for you.
What Is Predictive Maintenance?
Predictive maintenance is just what it sounds like: the science of predicting when maintenance will need to be performed on specific pieces of equipment. It can also be referred to as prescriptive maintenance or digital prescriptive maintenance. It’s usually driven by sensors that collect data about the specific pieces of equipment being maintained, after which analytics show when maintenance is most likely to be useful and least likely to cause problems.
In contrast, preventive maintenance is generally schedule-driven. Without knowing when equipment will break down, equipment managers schedule maintenance to prevent problems before they happen.
Both predictive and preventive maintenance are better than reactive maintenance, where equipment is maintained after there has already been a failure.
Why Not Just Use Tried-And-True Preventive Maintenance?
Preventive maintenance has been around for a long time and is relatively easy to manage. But while it’s better than no maintenance, it can be wasteful and even counterproductive.
Even idealized schedules will result in specific pieces of equipment being maintained more often than they need to be. Technicians will spend time disassembling equipment, lubricating or in other ways treating parts, replacing parts, and reassembling the equipment. During that process, they might break parts or cause other problems that wouldn’t have occurred if the equipment had been left alone longer. In addition, some pieces of equipment won’t get maintenance when they need it.
Predictive maintenance occurs when, and only when, it is most likely to be required for a given piece of equipment. When the predictions cause maintenance to be done earlier, it reduces the number of breakdowns that would happen with preventive maintenance; when it causes maintenance to be done later, it reduces the overall cost of maintenance and potential problems that would occur if excess maintenance had been done.
As sensor technology has become more affordable and predictive analytics technology easier to use, predictive maintenance has become more cost-effective and efficient.
Why is predictive maintenance important?
Most organizations take a reactive approach to maintenance: On average, 55 percent of maintenance is reactive, 31 percent is preventative, and just 12 percent is predictive.
Previously, the high cost of sensors and predictive analytics might have made it cheaper to fix a broken machine than to implement predictive maintenance. Now, the cost of proactively avoiding breakdowns can be much lower than waiting for breakdowns or even than performing preventive maintenance.
By switching from reactive or preventive maintenance to predictive maintenance, organizations can cut costs, improve efficiencies, and better perform their missions.
Are there emerging technologies that can make Predictive Maintenance easier?
The Internet of Things (IoT) provides many industries with the opportunity to manage maintenance more effectively. Devices used in an IoT scenario often have maintenance advantages over non-IoT-enabled devices, but when they’re combined with data management and analytics, these advantages can be realized across a plant, a supply chain, or an entire organization.
What are some predictive maintenance use cases?
Information Builders has a number of customers using data and analytics for predictive maintenance.
Cascades – Just-in-Time, Operations and Devices and IoT
The Tissue Group uses GE Proficy Plant Applications to monitor machinery and gather statistics about physical conditions. Sensors on the equipment gather data about temperature, humidity, and other operating conditions that affect the quality of the finished products. The IT department then uses integration technologies from Information Builders to obtain log files of this information and send those files to a central data center for processing.
Equipment sensors monitor variables within the production process and consolidate the results for offline analysis and this enables them to replace a machine before it fails!
Lipari Foods – Supply Chain and IoT
Lipari works with conveyor sensors and provides data for predictive maintenance. But beyond that Lipari Foods leverages IoT and analytics across from warehouse to deliver, this includes analytics for gauging truck speeds, container temperatures, idle time speeds, in an effort to optimization and analyze their entire supply chain.
Leading Defense Company – Leveraging IoT and Analytics across entire organization
A leading defense, aerospace, and security solutions company relies on data management, business intelligence (BI), and analytics technologies to help its customers better manage fleets and other large, complex assets. Data is gathered and integrated from supply-chain, design, maintenance, and other applications, as well as on-board systems, and then standardized to ensure compliance with ISO standards. Stakeholders can then interact with that information to proactively ensure that assets are being serviced and repaired in a way that optimizes performance and value.
Predictive Maintenance– Next Steps
Tune into this live webinar on Predictive Maintenance in the Public Sector. Learn how predictive maintenance improves how organization plans for the end of life of assets, which means better services for constituents and better-managed budgets.