Equipment Performance: If You Don’t Measure It, You Can’t Improve It
Most facilities know roughly how a piece of equipment is running. “Roughly” is the problem. It hides exactly where the losses are happening. Overall equipment effectiveness (OEE) is the standard metric for closing that gap. OEE is a single score built from three measurable factors that turns a vague sense of “running fine” into a number that points directly at where time and output are being lost.
Capturing OEE
OEE is calculated from three components multiplied together:
- Availability: Availability measures the ratio of actual run time to planned production time. This metric captures losses caused by unplanned downtime, changeovers, and equipment breakdowns.
- Performance: Performance measures how fast equipment operates compared to its designed speed. This metric captures losses caused by minor stops, slow cycles, and idling.
- Quality: Quality measures the ratio of good units produced to the total number of units started. This metric captures losses resulting from scrap, rework, and product defects.
OEE = Availability × Performance × Quality. A 100% score represents a theoretical perfect run: no downtime, full speed, zero defects. Multiplying the three factors instead of averaging them matters. A small loss in each one compounds into a much larger combined loss than any single number suggests.
The compounding effect
A machine running at 90% availability, 95% performance, and 98% quality looks healthy on every individual metric. Multiply them together, though, and the combined OEE comes out to 83.8%, which is noticeably lower than any of the three numbers in isolation. That gap is the recoverable capacity most facilities don’t know they have.
It’s also why tracking only one or two of the three factors (uptime alone, for instance) misses where the rest of the loss is hiding. A line that looks fine on availability can still be losing significant capacity to slow cycles or rework that nobody’s measuring.
What a baseline enables
A baseline OEE score for a piece of equipment makes future readings meaningful. Without one, “improved” or “declined” has no reference point to measure against. OEE scores can also be compared across shifts, lines, or sites of different sizes in a way raw output numbers can’t. It’s a useful property when prioritizing where to focus improvement efforts across a facility.
Breaking the score down by component directs attention to the right place. A maintenance team chasing speed improvements when the real loss is in availability is solving the wrong problem. The only way to know which factor is driving the loss is to track all three separately, not just the combined score.
Where the data comes from
The data sources are usually already available: downtime and changeover events from a CMMS or manual logs, actual cycle times against designed cycle time from the control system or manual timing, and quality or rework counts from inspection records. The collection method matters less than consistency. Measuring the same way over time is what makes the trend useful, even when the initial data collection is simple.
Pay attention to what OEE is telling you
A facility that doesn’t measure OEE isn’t necessarily running worse. It just has no way of knowing where its recoverable capacity is. The number isn’t the goal but the tool that turns “we should run better” into a specific, addressable loss category. Equipment that’s measured gets improved. Equipment that’s assumed to be fine usually isn’t, in ways nobody can point to.
