What is OEE? Availability, Performance and Quality explained

OEE, or Overall Equipment Effectiveness, is a single percentage that tells you how much of your planned production time is genuinely productive. It is the product of three factors: Availability, Performance, and Quality. An OEE of 100 percent means the equipment produced only good parts, at its fastest realistic speed, with no unplanned or planned stops, for the entire time it was scheduled to run. Every point below 100 percent points to a specific, nameable loss that someone can investigate and act on.

The value of OEE is not the number itself. It is the structure the number forces onto a messy reality. A machine that is down, a machine that runs slower than its design speed, and a machine that makes scrap are three very different problems. OEE separates those problems into three factors so that a plant manager can see, in one view, whether a low score is a maintenance issue, a speed issue, or a quality issue. This article defines the metric, walks through the formulas, works one complete numerical example, and finishes with the limits of the method, because a metric that is applied without knowing its limits will be misread.

Where OEE comes from

OEE was introduced by Seiichi Nakajima as part of Total Productive Maintenance (TPM), the equipment-management methodology developed in Japan in the 1970s and later published in English. In his 1988 book, Introduction to TPM, Nakajima described OEE as a way to surface the hidden losses that keep equipment from doing its job. The metric was later formalized as a key performance indicator in ISO 22400-2:2014, Automation systems and integration - Key performance indicators (KPIs) for manufacturing operations management - Part 2: Definitions and descriptions. That standard gives OEE a formal place in the manufacturing operations management (MOM) layer, alongside a set of defined time elements such as planned busy time, actual production time, and good quantity.

There is a meaningful difference between the two traditions. The TPM tradition, following Nakajima, expresses OEE as the product of Availability, Performance, and Quality, and maps each factor to two of the Six Big Losses. The ISO 22400-2 standard defines OEE within a formal time model that uses different element names. Peer-reviewed work, notably Schiraldi and Varisco (2020) in Computers and Industrial Engineering, has shown that the ISO standard contains more than one OEE formulation and that these are not automatically consistent with the TPM model. The practical lesson is that before two teams, or two systems, compare OEE numbers, they must agree on which formulation they are using. We return to this point in the limitations section.

The three factors, defined

The preferred way to calculate OEE is to build it from three factors. Each factor answers one question about the equipment.

Availability answers: was the machine running when it was supposed to be? It is the ratio of Run Time to Planned Production Time.

Performance answers: when the machine was running, was it running as fast as it should? It is the ratio of the fastest possible time to make the parts, to the time actually taken.

Quality answers: of everything the machine produced, how much was good on the first pass? It is the ratio of Good Count to Total Count.

Before the formulas, four time and count terms need to be fixed, because small differences in how these are defined change the whole result.

Planned Production Time is the time the equipment is expected to produce. It starts from shift length and subtracts everything with no intention of running: breaks, lunches, scheduled maintenance, and periods with no orders. It is the denominator of the Availability factor.

Run Time is Planned Production Time minus Stop Time. Stop Time covers both unplanned stops, such as breakdowns and material shortages, and planned stops, such as changeovers and setups. Both kinds count, because both are time that could have been used for production.

Ideal Cycle Time is the fastest time to make one part under optimal conditions. It is also called the theoretical or nameplate cycle time. It is the reference speed against which Performance is measured.

Total Count and Good Count are the number of parts produced and the number of those parts that passed quality the first time, with no rework counted as good. OEE treats quality the way First Pass Yield does: a part that needs rework is not a good part for this calculation.

The formulas

With those terms fixed, the calculation is straightforward.

Availability = Run Time / Planned Production Time

Performance = (Ideal Cycle Time × Total Count) / Run Time

Quality = Good Count / Total Count

OEE = Availability × Performance × Quality

Multiplying the three factors together and reducing to the simplest terms gives the same result in a single step:

OEE = (Good Count × Ideal Cycle Time) / Planned Production Time

The single-step form is mathematically identical, but it hides the three factors. The preferred three-factor form is more useful, because a 75 percent OEE built from 90 percent Availability, 90 percent Performance, and 93 percent Quality points the improvement effort in a different direction than a 75 percent OEE built from 80 percent, 98 percent, and 96 percent. The single number records where you are; the three factors tell you where to look next.

A few practical notes on each factor. Availability treats a two-hour changeover the same way it treats a two-hour breakdown, which is deliberate: both are time the equipment was not producing. Performance should never exceed 100 percent in a correctly set-up system; if it does, the Ideal Cycle Time is set too slow and should be corrected, not celebrated. Quality measures only first-pass output, so rework and scrap both lower it.

The Six Big Losses

Nakajima tied each OEE factor to two concrete loss categories, a framework known as the Six Big Losses. The framework matters because it converts a score into a work list.

Availability losses:

• Equipment Failure. Unplanned stops such as breakdowns, tooling failures, and unplanned maintenance.
• Setup and Adjustments. Planned stops such as changeovers, setups, cleaning, and warmup.

Performance losses:

• Idling and Minor Stops. Short stops, usually under a few minutes, that an operator clears, such as misfeeds and jams.
• Reduced Speed. Running slower than Ideal Cycle Time, from worn parts, poor materials, or deliberate slowdown.

Quality losses:

• Process Defects. Scrap and rework produced during stable, steady-state production.
• Reduced Yield. Scrap and rework produced from startup until stable production is reached, most often right after a changeover.

The Six Big Losses give each factor a mechanism. When Availability is low, the plant looks at breakdowns and changeover time. When Performance is low, it looks at small stops and running speed. When Quality is low, it looks at steady-state defects and startup yield. Improvement becomes a sequence of named, bounded problems instead of a vague instruction to "do better."

A worked example

The assumptions are laid out in full so the calculation can be reproduced and checked.

Assumptions for one shift:

• Shift length: 8 hours = 480 minutes
• Planned non-production time: two 15-minute breaks and 30 minutes of scheduled maintenance = 60 minutes
• Unplanned stop time (breakdowns plus a changeover): 50 minutes
• Ideal Cycle Time (design speed): 2.0 seconds per part
• Total parts produced: 9,800
• Rejected parts: 300

Step 1. Planned Production Time = 480 - 60 = 420 minutes.

Step 2. Run Time = 420 - 50 = 370 minutes.

Step 3. Availability = 370 / 420 = 0.881, or 88.1 percent.

Step 4. Performance = (2.0 seconds × 9,800) / (370 minutes × 60 seconds) = 19,600 / 22,200 = 0.883, or 88.3 percent.

Step 5. Quality = (9,800 - 300) / 9,800 = 9,500 / 9,800 = 0.969, or 96.9 percent.

Step 6. OEE = 0.881 × 0.883 × 0.969 = 0.754, or 75.4 percent.

The single-step check gives the same answer: OEE = (9,500 good parts × 2.0 seconds) / (420 minutes × 60 seconds) = 19,000 / 25,200 = 0.754, or 75.4 percent.

Reading the result: the plant lost about 12 percent of planned time to stops, about 12 percent of running time to speed loss, and about 3 percent of output to defects. The three factors multiply to show that only about three quarters of the scheduled production time ended up as good output. The next move is not to argue about the number but to pick the largest loss, Availability in this case, and ask which of its two sub-losses, breakdowns or changeovers, is doing the most damage.

What OEE does not measure

OEE is precise about one thing, planned production time, and silent about everything outside it. Several common misuses follow from that.

OEE is not a benchmark between machines or plants. Two lines can report the same OEE for different reasons, and a line that runs one product will naturally score higher than a line that runs ten and eats changeover time. Comparing raw OEE across dissimilar equipment, or aggregating it to department or plant level, produces numbers that look comparable but are not.

The 85 percent "world-class" figure is context, not law. The familiar target of 90 percent Availability, 95 percent Performance, and 99 percent Quality, which multiplies to roughly 85 percent, comes from Nakajima's experience in Japanese automotive plants in the 1970s. It is a useful reference, but it is not a universal benchmark. Many plants run closer to 60 percent, and chasing a number borrowed from a different industry, era, and product mix can do more harm than good. The useful target is a stretch goal for your own process, not someone else's number.

OEE ignores time the equipment was never scheduled to run. Plant shutdowns, lack of orders, and calendar time outside the shift are excluded by design. If a manager wants to see effectiveness against all available time, the related metric is TEEP, Total Effective Equipment Performance, which multiplies OEE by a Loading factor. Confusing the two flatters the OEE number.

OEE is not a measure of people. It measures equipment losses. Using it to score operator performance invites people to game the data, for example by hiding small stops or inflating the good count, which destroys the very signal the metric exists to produce.

OEE depends on definitions. The choice of what counts as a "stop" versus a "minor stop," typically a recording threshold of one to five minutes, and the choice of Ideal Cycle Time, both move the result. Two systems that appear to report the same metric can disagree because they set these thresholds differently. The same caveat applies to the ISO 22400-2 formulation, which as noted above is not automatically identical to the TPM formulation.

OEE can be raised in ways that hurt the business. Reducing the number of changeovers, and thus running larger batches, raises Availability and therefore OEE, but it also raises inventory and works against lean flow. A high OEE achieved by overproducing parts nobody needs is not a win. The goal is to expose and remove losses, not to maximize the number at any cost.

Where to start

A plant that is new to OEE should begin smaller than it thinks it needs to. Pick one critical asset, agree on the definitions (Planned Production Time, the stop threshold, and Ideal Cycle Time) in writing, and capture data manually or with a simple system for a few weeks before automating anything. The first goal is a stable, believable baseline, not a perfect measurement system. From that baseline, the Six Big Losses become the improvement backlog: reduce the biggest loss first, re-measure, and repeat. When the number is stable and the team trusts it, then the question of scaling measurement across the plant, or connecting it to an MES or other system, becomes worth answering.

Sources

1. ISO 22400-2:2014, Automation systems and integration - Key performance indicators (KPIs) for manufacturing operations management - Part 2: Definitions and descriptions. International Organization for Standardization. https://www.iso.org/standard/54497.html
2. Nakajima, Seiichi (1988). Introduction to TPM: Total Productive Maintenance. Productivity Press. ISBN 978-0-915299-23-2.
3. Schiraldi, Massimiliano M., & Varisco, Martina (2020). Overall Equipment Effectiveness: consistency of ISO standard with literature. Computers & Industrial Engineering, 145, 106518. doi:10.1016/j.cie.2020.106518.
4. Vorne Industries. OEE Calculation: Definitions, Formulas, and Examples. https://www.oee.com/calculating-oee/ ; OEE Factors. https://www.oee.com/oee-factors/ ; Six Big Losses in Manufacturing. https://www.oee.com/oee-six-big-losses/ ; World-Class OEE. https://www.oee.com/world-class-oee/

Related reading

• How to calculate OEE without hiding production losses
• OEE vs TEEP vs capacity utilization
• Six common OEE measurement errors