How to Calculate OEE Correctly in a Real Factory

OEE looks simple — one percentage that says how well a machine is really running. But on most floors it's calculated inconsistently: breaks included on one line and excluded on another, "downtime" defined differently by each supervisor, ideal rates that nobody has revisited in years. The result is a number that can't be compared shift-to-shift or line-to-line — and that people quietly stop trusting.

What OEE actually is

OEE (Overall Equipment Effectiveness) is the product of three ratios:

• Availability = run time ÷ planned production time. It captures time lost to stops.
• Performance = actual output ÷ what output would have been at the ideal rate. It captures slow cycles and small stops.
• Quality = good units ÷ total units produced. It captures scrap and rework.

Each is a fraction between 0 and 1; multiply the three for OEE. As a purely illustrative example: if a line is available 90% of its planned time, runs at 95% of its ideal rate, and produces 99% good parts, OEE ≈ 0.90 × 0.95 × 0.99 ≈ 0.85. (Those numbers are hypothetical, just to show the arithmetic — not a benchmark or a result to aim at.)

The part that actually matters

Every one of those three ratios depends on a definition you choose: what counts as planned production time, what's "downtime" versus a "planned stop," what the ideal rate is for a given product, and where you count "good." Change a definition and the number changes. So calculating OEE "correctly" is less about the formula — which is trivial — and more about agreeing on consistent definitions and applying them the same way everywhere.

Where it applies

OEE is most straightforward on discrete manufacturing — CNC machining, presses, assembly, packaging lines — where cycles and counts are well-defined. Continuous and batch processes can use it too, but the state and counting definitions need more care.

Common mistakes teams make

• Inconsistent "planned production time." Whether breaks, planned maintenance, and no-demand time are in or out is one major reason two lines' OEE can become difficult to compare.
• Blurring planned and unplanned stops. A scheduled changeover and a surprise breakdown are not the same loss, and lumping them hides the actionable one.
• A stale or guessed ideal rate. If "ideal" isn't grounded per product and machine, Performance is fiction.
• Counting quality at the wrong point. Before or after rework gives very different numbers.
• Manual data entry. Operator-logged downtime can be slow, incomplete, and difficult to audit later.

What makes an OEE calculation trustworthy

• Agreed, documented state definitions — running, planned stop, unplanned stop, idle, setup, inspection — owned by you.
• A documented, product-specific ideal rate, with a clear process for updating it when products, tooling, or process conditions change.
• Clear good-versus-total counting rules.
• Automated, timestamped capture from the machine, not from memory.
• The same rules applied on every line, every shift.

How Elpis approaches it

EdgeConnect collects the machine signals; EREMOS V2 computes OEE from explicit machine states against your own definitions, not a vendor preset, on traceable, timestamped signals — and applies those rules consistently across lines and sites. The software doesn't decide what "running" means for your process; you define it, and it applies it the same way every time. More in the Operational Intelligence capability overview.

Where to start

Write down your state definitions and your ideal rates before you automate anything — the calculation is only as good as those.

Elpis IT Solutions builds an Industrial Intelligence Ecosystem — from shop-floor signal to enterprise decision. Operating across India and the Middle East.