⚙️ OEE Explained: Availability × Performance × Quality

What Is OEE?

Overall Equipment Effectiveness (OEE) is the gold-standard metric for measuring how effectively a manufacturing operation uses its equipment. It boils a complex production picture down to a single percentage that answers one question: of the time we planned to produce good parts at full speed, what fraction did we actually achieve?

OEE is the backbone of Total Productive Maintenance (TPM) and a common headline KPI on factory dashboards. Its power lies in its structure — it is the product of three intuitive factors, each tied to a distinct category of loss. You can run your own numbers on the OEE calculator.

The Three Factors

OEE = Availability × Performance × Quality

• Availability = Run Time ÷ Planned Production Time. It captures losses from breakdowns, setups, and unplanned stops.
• Performance = (Ideal Cycle Time × Total Count) ÷ Run Time. It captures speed losses — running slower than the equipment's rated speed, plus minor stops and idling.
• Quality = Good Count ÷ Total Count. It captures defects, rework, and startup scrap.

Each factor is a fraction between 0 and 1. Multiplying them gives an honest, compounded view: a line that is 90% available, runs at 95% of rated speed, and produces 99% good parts is not 90%+ effective — it is 0.90 × 0.95 × 0.99 = 84.6% OEE.

A Worked Example

Consider one 8-hour shift on a packaging line:

• Shift length: 8 hours = 480 minutes
• Planned breaks: 30 minutes → Planned Production Time = 450 minutes
• Unplanned downtime (breakdowns + changeover overrun): 45 minutes → Run Time = 405 minutes
• Ideal cycle time: 1.0 second per unit (60 units/minute)
• Total count produced: 19,440 units
• Reject count: 440 units → Good Count = 19,000 units

Availability = 405 ÷ 450 = 0.900 (90.0%)

Performance = (1.0 s × 19,440 units) ÷ (405 min × 60 s) = 19,440 ÷ 24,300 = 0.800 (80.0%)

Quality = 19,000 ÷ 19,440 = 0.9774 (97.74%)

OEE = 0.900 × 0.800 × 0.9774 = 0.7037 ≈ 70.4%

This line is running at about 70% OEE — a solid but improvable result. The biggest opportunity here is clearly performance (80%): the equipment could produce 24,300 units in the available run time but made only 19,440, suggesting speed loss and minor stops worth investigating.

The Six Big Losses

Each OEE factor maps to two of the classic "six big losses." Mapping a low factor to its losses tells you exactly where to focus improvement effort:

OEE Factor	Big Loss	Typical Cause
Availability	Equipment failure	Breakdowns, unplanned maintenance
Availability	Setup & adjustment	Changeovers, tooling changes, warm-up
Performance	Idling & minor stops	Jams, blocked sensors, brief stoppages
Performance	Reduced speed	Running below rated cycle time
Quality	Process defects	Scrap and rework during steady-state
Quality	Startup & yield losses	Defects during warm-up or after changeover

What Is World-Class OEE?

Benchmarks for discrete manufacturing are widely cited:

• 100% — perfect production: only good parts, at full speed, with zero stops.
• 85% — world-class for many discrete manufacturers; a long-term goal for most.
• 60% — fairly typical; substantial room for improvement.
• 40% — common for plants just starting to measure OEE; usually low-hanging fruit everywhere.

The world-class 85% target is typically broken down as roughly 90% availability × 95% performance × 99% quality. Note that chasing a single high number can mislead — a plant can inflate availability by producing slowly (no stops) while quietly destroying performance. Always look at the three factors separately.

OEE, TEEP, and Common Mistakes

Standard OEE measures against planned production time. If you want to capture schedule loss (time the equipment could have run but wasn't scheduled), use TEEP = OEE × Utilization, which measures against all calendar time. Common mistakes when implementing OEE include:

• Using a "nameplate" cycle time that is faster than reality, which understates performance unfairly.
• Excluding minor stops because they're hard to log — they often add up to the biggest loss.
• Comparing OEE across dissimilar machines as if the number were absolute; OEE is best used as a trend on one process.
• Treating OEE as a worker scorecard rather than a process-improvement compass, which discourages honest data.