How to improve OEE with predictive maintenance from AiSight

Overall equipment effectiveness (OEE) may be the most important KPI in industry. Improving OEE entails sweeping improvements in efficiency and product quality; it benefits stakeholders along every part of the value chain. Anyone in production, from continuous improvement managers and reliability engineers, to production managers, to maintenance technicians, wants improved OEE.

And using predictive maintenance, such as AiSight's solution, will drastically improve OEE.

This article explains how predictive maintenance improves OEE. We’ll take a deeper look at OEE and the impediments to its improvement, before getting into predictive maintenance and its advantages. With a clear understanding of OEE in hand, the power of predictive maintenance to improve OEE becomes clear—this is the tool you need to improve OEE in your industry.

What is OEE?

Overall equipment effectiveness is a measurement of relative productivity. This metric compares actual equipment performance to equipment capacity, expressing the rating as a percentage—an OEE score of 100% means your equipment could not operate more efficiently.

OEE provides a comprehensive, three-dimensional picture of equipment operations. The three dimensions of OEE are availability, performance, and quality. When generating an OEE score, we measure each and assign each a percentage score.

• Availability is a measurement of uptime—the percentage of scheduled time during which equipment actually operates. Equipment with 100% availability suffers no unplanned downtime.
• Performance measures the speed of output as a percentage of equipment’s designed output. 100% performance indicates equipment working at capacity.
• Quality measures the quantity of produced units that meet quality control standards, compared to the total number of units started. 100% quality indicates production with no defective output.

We calculate the final OEE score by multiplying each of these factors together. A production line with 90% availability, 90% performance, and 90% quality thus has an OEE score of 73%. Because we're multiplying three dimensions, like calculating the volume of a cube, we can visualize this OEE score as a bin that's roughly ¾ full.

Why monitor overall equipment effectiveness?

OEE was developed as an indicator in Seiichi Nakajima’s total productive management (TPM) method of asset management. Nakajima built TPM on eight pillars of proactive and preventative maintenance practices: focused improvement, autonomous maintenance, planned maintenance, quality maintenance, development maintenance, education and training, administrative maintenance, and safety and environmental maintenance.

TPM engages stakeholders in small-scale, semi-autonomous activities that in aggregate affect continuous improvement in equipment effectiveness. OEE is the indicator of where to direct those activities.

Using OEE, we can identify problems in our processes, then deliberately and proactively address them. While OEE is a measure of effectiveness, as a comparative measure it contains implicit calculations of limits to effectiveness—the things we want to eliminate in our plants.

What limits OEE?

In measuring availability, performance, and quality to calculate OEE, we need to monitor the things that reduce availability, performance, and quality.

Nakajima identified “Six Big Losses” that limit OEE—two for each of the three dimensions contributing to OEE.

• Losses to availability: equipment failures and planned downtime
• Losses to performance: small stops and reduced speed
• Losses to quality: rejects and reduced yield.

If we consider a hypothetical plant, we can see how each of these losses limit OEE.

In our hypothetical plant, we make something very important: plastic army men. It’s not a complicated process, but our equipment is old and ineffective. We aren’t working with anywhere near 100% OEE.

On a given day, the problems start as soon as production is supposed to. We can count on starting every eight-hour shift by spending 30 minutes kicking machines until they work—this counts as planned downtime. Despite this daily interval of planned maintenance, we can also count on something breaking down, robbing us of another two hours of equipment availability. Our equipment availability stands at an abysmal 69%.

Back in the day, this plant used to churn out 500 American heroes every hour. But nothing runs like it used to. Most equipment only runs as well as it does with a lubrication break between reach run, and material jams in the bagging machine require manual intervention. In any given hour we’ll make about 380 units, for a 76% performance score.

Finally, this old equipment produces poor-quality soldiers. The plastic melter takes its time heating up and the first platoon of the shift comes out looking like green Swiss cheese. And something’s wrong with a mold for the prone soldier—about half of them need a medic right off the line. Luckily, standards for plastic army men aren’t high. Even with those losses, we can claim 90% quality.

Thanks to the many issues in our hypothetical plant, we're looking at an overall equipment effectiveness score of only 47%.

Luckily, there are ways to improve OEE—even in plants with far better OEE scores than our hypothetical case! Improving OEE calls for the kind of proactive maintenance recommended under total productive management. Taking preventative measures puts us ahead of problems before they can reduce OEE, and gives us the opportunity to improve OEE. We just need to know what we have to prevent.

For that, there’s predictive maintenance.

What is predictive maintenance?

Predictive maintenance describes data-driven maintenance practices, informed by automated machine monitoring, sensor readings, and data analysis. AiSight's predictive maintenance solution mounts sensors nodes on industrial equipment, then constantly monitors that equipment with cloud-based artificial intelligence. Using machine-learning algorithms and vibration analysis, our solution identifies the optimal parameters for machine operations, and alerts you when machines operate outside of those parameters—allowing you to plan maintenance months before any failure.

Predictive maintenance is part of a proactive approach to eliminating machine faults and breakdowns before they become problems.  Compared to preventative maintenance, which goes on whether or not there's a problem, or reactive maintenance, which can only occur after a breakdown, the advantages of predictive maintenance are clear.

By deploying AiSight's predictive maintenance solution, you’ll know when your machines will break down, why they’ll break down, and what you can do about it—long before anything actually goes wrong.

This makes predictive maintenance a potent way to improve OEE.

Learn more about the advantages of predictive maintenance.

How to improve OEE with predictive maintenance

At AiSight, we take ending unplanned downtime seriously. And our predictive maintenance solution provides the information you need to prevent equipment failures. That alone can eliminate one of the 6 Big Losses, thus improving OEE. But that’s far from all that predictive maintenance can do to improve OEE.

How predictive maintenance improves equipment availability

By predicting machine malfunctions months before they happen, AiSight’s predictive maintenance solution provides the information you need to plan maintenance ahead and perform proactive machine maintenance.

Predictive maintenance, therefore, prevents equipment failures, improves equipment availability, and improves OEE.

But AiSight’s predictive maintenance solution also provides the information you need to reduce planned downtime too.

By providing a clear picture of machine health, predictive maintenance eliminates unnecessary preventative maintenance and decreases turnaround time during planned maintenance intervals. This contributes further to the big gains in equipment availability possible by using predictive maintenance—and further improves OEE

How predictive maintenance improves equipment performance

Healthy machines perform. By using predictive maintenance to monitor machine health, it’s possible to provide the maintenance they need to operate at maximum capacity. Equipment performance need never suffer.

Even better, AiSight’s predictive maintenance solution provides monitoring and analysis capable of identifying many process errors that necessitate small stops. It’s therefore possible to keep equipment running—and performing.

How predictive maintenance improves quality

By eliminating breakdowns and process errors, it’s possible to reduce the material waste that comes with it. This means eliminating scrap that occurs during non-optimal machine operation—an improvement in production quality.

Properly functioning machines also operate with improved regularity and fewer production irregularities. By monitoring and ensuring machine health with predictive maintenance, it’s possible to reduce quality control issues.

With predictive maintenance facilitating improvements across the board in equipment availability, performance, and production quality, we come to the obvious conclusion: predictive maintenance improves OEE.

AiSight’s predictive maintenance solution is the tool you need to improve OEE in your plant. Getting started is easy. We monitor almost any type of rotating equipment, in almost any industry. We provide machine insights on day one, and full support as long as you use our solution. Let’s get started.

Get a free consultation