OEE vs efficiency vs utilization

OEE, efficiency and utilization all sound like "how well is the machine doing?", so they get quoted interchangeably in the same meeting and quietly mean different things. They measure against different baselines, so a line can score high on one and poorly on another. Here's what each really tells you, and why OEE is the one that catches losses the other two hide.

Utilization: was the machine busy?

Utilization is the share of available time the machine was running. It answers one question: was it on?

Its blind spot is everything that happens while it runs. A machine running flat out but making half-speed scrap can post 95% utilization: it was busy, just not productive. Utilization sees no difference between a good part and a bad one, or full speed and half speed. The "available time" you choose also changes the number completely (all 24 hours? only staffed shifts?), so utilization figures are easy to dress up.

Efficiency: did it hit the standard?

Efficiency compares actual output to an expected standard, usually a rate or a labour standard. It answers: did we hit the number we said we would?

The catch is the word "standard." Efficiency is only as honest as the target it's measured against. If the standard rate is set conservatively, you can run at 110% efficiency and still be leaving a third of the line's real capacity on the floor. Efficiency tells you how you did versus expectation, not how you did versus what was possible.

OEE: was production effective, all three ways at once?

OEE rolls the three questions the others each answer only partly (was it running, was it fast, was it good) into one figure, measured against perfect production:

Because it multiplies all three, OEE can't be gamed the way utilization (ignores speed and quality) or efficiency (depends on a soft standard) can. A line that's 95% utilized but running slow and scrapping parts will show a high utilization and a mediocre OEE, and the OEE is the one telling you the truth.

Side by side

• Utilization measures against available time. Catches idle time only; blind to speed and quality.
• Efficiency measures against a standard you set. Catches rate vs target; only as honest as the standard.
• OEE measures against perfect production. Catches downtime, slow running and defects together; the hardest to fake.

None of them is wrong; they answer different questions. But if you're going to track one number to drive improvement, make it OEE, because it's the only one that points you at all three kinds of loss. Just don't confuse a great utilization figure for a productive line. For the related family of effectiveness ratios, see OEE vs TEEP vs NEE, and for what a strong score looks like, what is a good OEE.