October 19, 2022
By: Daniel Voelp
Leveraging Real-Time Data for Assessing Process Plant OEE
The concept of Overall Equipment Effectiveness (OEE), a key manufacturing measurement for decades, has gained new momentum as plants are going digital. Processing plant managers and researchers are increasingly embracing OEE as a tool for achieving lean production goals, boosting productivity, and cutting waste. Processing plant managers and researchers have touted the method, today enhanced by AI, as a solution for improving productivity and cutting costs.
As digital transformation gathers pace, plant assets are becoming more complex and more costly, making OEE even more important to help plant managers gain a quick snapshot of the plant’s efficiency and measure its improvement. Fortunately, the same digital transformation is also enabling plants to implement advanced technologies like artificial intelligence (AI) and machine learning (ML) to make OEE metrics faster and more reliable by basing them on real-time data.
What’s Behind the Letters O, E, and E? Making the Right Calculation
At the heart of it, OEE is a percentage that shows the plant’s actual output, divided by the maximum potential output that the plant could theoretically achieve. To reach that percentage, we use three individual factors: availability, performance, and quality. This way, you can consider the quality of the product as well as how much product the plant can produce.
How to calculate OEE?
The widely accepted formula for calculating OEE is:
OEE = Availability × Performance × Quality
Or to be more specific:
100 x Actual (Availability x Performance x Quality) / Theoretical Maximum (Availability x Performance x Quality)
Here’s how to understand the metrics for process manufacturing plants:
Availability tracks how much time the plant or equipment is running. It considers downtime (whether for maintenance or due to breakdowns), and when downtime occurs. In theory, a piece of equipment could be available 24/7 without any breaks. In practice, sometimes it’s not working. Availability measures how long it’s actually available compared with how long it could be available.
Availability = (actual running time/time available for running) x 100
Performance calculates how fast the equipment or the plant is performing, compared to its maximum possible speed. It includes slow production speed, pauses between cycles, and any time that equipment has to wait because the previous stage of production took longer to complete and the product isn’t ready yet.
Performance = (actual hourly performance rate/average of top 6 hourly rates over time) x 100
Quality is perhaps the most complicated metric because any given stage in process manufacturing might not produce discrete products that can be measured and tested for quality. In fact, the “product” of most of the stages of production are intermediates, and you won’t find a finished product that can be quality-tested until the end of the entire process. This means that acceptable quality metrics are often a range, rather than a single figure. It also means that different verticals can use different metrics for their quality formula.
For example, this formula from LearnQCTools considers “good” quality to be products that weren’t rejected or sent back for total rework, which might still include some lower-quality product.
Quality = ((total production-total rejected)/total production) * 100
But the formula used by Altus takes standard quality items as its measure for “good” quality, which might mean it excludes product that LearnQCTools would include:
Quality = (Quantity of Standard Items / Total Quantity Produced) * 100%
Constant Real-Time Updates Boost OEE
As you can see, calculating OEE relies on first carrying out accurate and reliable calculations for Availability, Performance, and Quality, and those in turn depend on fast, trustworthy data. This is where real-time data comes into play.
AI and ML analytics like Precognize’s SAM GUARD software can crunch real time data from Industrial Internet of Things (IIoT) devices and plant sensors swiftly enough to keep your calculations updated, giving you a trackable, real-time overview of plant operations. Up-to-date insights from SAM GUARD into the availability, performance and quality of a process allow you to understand how well you’re meeting your production expectations, so you can predict and strategize your capabilities and needs for the future.
What’s more, SAM GUARD’s AI-powered virtual sensors go beyond the standard measurements like current, speed, weight, and temperature which are used to calculate availability and production levels. This data identifies the quality level of intermediates and outputs in real-time, enabling OEE to be measured at any moment instead of waiting for timed lab tests.
In addition to helping with OEE measurements, SAM GUARD’s real-time data analysis actively boosts OEE numbers. The constant flow of accurate data into SAM GUARD enables it to issue relevant alerts the moment any anomaly arises, notifying employees about nascent issues within the plant. As a result, plant employees can quickly zero in on any problem affecting availability, performance, or quality, which would in turn drag down OEE. Early alerts allow managers to change processes sooner, or pause production to remediate the issue while it’s still relatively minor and be fixed in a shorter space of time, helping keep OEE as high as possible.
Real-Time Data Brings OEE to New Heights
OEE has come a long way since it was first introduced by Japanese manufacturing industry pioneer Seiichi Nakajima in the 1960s to assess manufacturing operation effectiveness. While data was once collected manually and calculated for a single piece of equipment at a time, more modern OEE calculations factor in various machines along the processing chain. Today, OEE is further optimized by automated solutions like Precognize’s SAM GUARD, which track equipment and processes in real time across the entire plant.
By identifying inefficiencies, Precognize’s OEE software can help improve every aspect that affects OEE, and help you come closer to achieving zero delays, breakdowns, and impediments to production.
Would you like to learn more? Get a Demo