Published March 24, 2026 · 8 min read
OEE stands for Overall Equipment Effectiveness. It is the single most widely used metric for measuring manufacturing productivity. OEE tells you what percentage of planned production time is truly productive — meaning you are making good parts, at full speed, with no downtime.
The metric was originally developed by Seiichi Nakajima as part of Total Productive Maintenance (TPM) in the 1960s. It has since been formalized in the international standard ISO 22400-2 (Manufacturing operations management — Key performance indicators), which defines OEE and its sub-components for discrete, continuous, and batch manufacturing.
An OEE score of 100% means you are manufacturing only good parts, as fast as possible, with zero downtime. In practice, no plant runs at 100%. But understanding how far you are from it — and which component is pulling you down — is what makes OEE so powerful.
OEE is the product of three independent factors:
OEE = Availability × Performance × Quality
Each factor is a percentage between 0% and 100%. Because they multiply together, OEE is always less than or equal to the lowest individual factor. This is why even "small" losses in each category compound into significant overall losses.
For example, if each factor is 90%, your OEE is not 90% — it is 0.90 × 0.90 × 0.90 = 72.9%. That compounding effect is exactly why OEE exposes hidden capacity that other metrics miss.
Availability measures the percentage of planned production time that the equipment was actually running. It accounts for all events that stop production for an appreciable length of time.
Availability = Run Time ÷ Planned Production Time
Planned Production Time is the total shift time minus any planned stops. Planned stops include scheduled breaks, planned maintenance windows, and periods where there are no orders to run. These are excluded because the equipment was never intended to produce during those periods.
Run Time is the Planned Production Time minus unplanned stops. Unplanned stops include:
Note that some organizations classify changeover as planned downtime. The important thing is consistency: pick a definition and stick with it across all machines so comparisons are meaningful.
Performance measures whether the equipment is running at its maximum designed speed. It captures anything that causes the process to run slower than the theoretical maximum.
Performance = (Ideal Cycle Time × Total Count) ÷ Run Time
Equivalently, Performance = (Total Count / Run Time) / Ideal Run Rate, where Ideal Run Rate is the maximum number of parts per minute the equipment can produce.
Ideal Cycle Time is the theoretical minimum time to produce one part. It should come from the equipment manufacturer's specification or from a validated time study — not from an average of recent production. Using an average will mask speed losses.
Performance losses include:
Performance can exceed 100% if the machine runs faster than the documented ideal cycle time. This usually means the ideal cycle time is set too conservatively and should be updated.
Quality measures the proportion of produced parts that meet specifications on the first pass, without any rework.
Quality = Good Count ÷ Total Count
Quality losses include:
A common mistake is counting reworked parts as "good" because they eventually ship. In OEE, rework is a quality loss because the machine time spent on the first (failed) attempt was wasted.
Let's walk through a complete OEE calculation for a plastics extrusion line during a single 8-hour shift.
| Parameter | Value |
|---|---|
| Total shift time | 480 minutes (8 hours) |
| Planned downtime (breaks) | 30 minutes |
| Unplanned downtime (breakdown + changeover) | 45 minutes |
| Ideal cycle time | 1 part per minute |
| Total parts produced | 340 parts |
| Rejected parts | 15 parts |
Planned Production Time = Total shift time − Planned downtime = 480 − 30 = 450 minutes
Run Time = Planned Production Time − Unplanned downtime = 450 − 45 = 405 minutes
Availability = 405 ÷ 450 = 90.0%
At an ideal cycle time of 1 part per minute, the line should have produced 405 parts in 405 minutes of run time. It actually produced 340.
Performance = (1 min/part × 340 parts) ÷ 405 min = 340 ÷ 405 = 84.0%
This means 16% of the available run time was lost to minor stops, reduced speed, or slow cycles. That is 65 minutes of hidden capacity — time the machine was technically "running" but not producing at full rate.
Good Count = Total parts − Rejected parts = 340 − 15 = 325 parts
Quality = 325 ÷ 340 = 95.6%
| Component | Value |
|---|---|
| Availability | 90.0% |
| Performance | 84.0% |
| Quality | 95.6% |
| OEE | 72.3% |
OEE = 0.90 × 0.84 × 0.956 = 72.3%
Out of 450 minutes of planned production time, only 325 minutes' worth of good product was actually created. That means 125 minutes — over two hours — of potential capacity was lost to a combination of downtime, speed loss, and defects.
The widely cited benchmark for world-class OEE is 85%. This is generally broken down as:
These are targets, not guarantees. Most manufacturing plants that first begin measuring OEE discover they are operating somewhere between 40% and 60%. A plant consistently hitting 85% has very tight control over downtime, is running close to nameplate speed, and has near-zero defects.
It is important to note that the 85% benchmark assumes discrete manufacturing with relatively stable, repetitive processes. For job shops with frequent changeovers, batch chemical processes, or custom fabrication, the target may be different. What matters more than the absolute number is the trend over time and the ability to identify which of the three factors is the biggest constraint.
For context, studies from the manufacturing consulting industry consistently find that the average plant operates at roughly 60% OEE. That means 40% of available production capacity is being lost. Even a 5-percentage-point improvement in OEE can translate to hundreds of thousands of dollars in additional output per year on a single production line.
If you include scheduled breaks or planned maintenance in the denominator, your Availability will be artificially low and you will not be able to distinguish between controllable losses and intentional stops. Planned downtime should be subtracted from total time before calculating Availability.
If you calculate Performance using the average observed cycle time, your Performance will always be close to 100%, which defeats the purpose. The ideal cycle time must reflect the machine's maximum demonstrated capability, not its recent average. If a CNC mill can cut a part in 3.2 minutes under optimal conditions, that is the ideal cycle time — even if it currently averages 4.1 minutes.
Parts that require rework consumed machine time and raw material on the first pass. Even if they are eventually corrected and shipped, the Quality factor should only count parts that were right the first time. Failing to account for rework inflates Quality and hides a real source of waste.
If one production line counts changeover as planned downtime and another counts it as unplanned, their OEE numbers are not comparable. Standardize your definitions across all equipment before benchmarking machines against each other.
OEE is a diagnostic tool, not a trophy. The value is in the breakdown: knowing that your biggest loss is Availability (downtime) points you toward maintenance improvements, while a Performance problem points toward speed optimization or minor-stop reduction. Collecting OEE data without a process for investigating and addressing the losses is waste in itself.
Relying on operators to manually log downtime events, part counts, and reject reasons introduces delays and inaccuracies. Operators tend to round numbers, forget short stops, and categorize downtime inconsistently. Automated data collection from the machine's PLC or sensors eliminates these issues and provides real-time visibility instead of end-of-shift reports.
PulseMQ connects directly to your PLCs via MQTT and Sparkplug B, pulling live signals for machine state, part counts, and quality events. OEE is calculated automatically in real time with no manual data entry, no spreadsheets, and no end-of-shift guesswork. Each component — Availability, Performance, and Quality — is broken down on your dashboard so you can see exactly where losses are occurring and act on them immediately.
The platform supports discrete, continuous, and batch manufacturing processes in accordance with ISA-88, ISA-106, and ISO 22400-2 standards. Learn more about the PulseMQ platform.
Stop guessing. Connect your machines and get accurate, automated OEE calculations on every line, every shift, every day.
Contact Sales