Machine tools · OEE reality check
Not because anyone is lying — because the small stops nobody logs and the speed loss nobody measures quietly drop out of the maths. The figures shown are an example shift. Put your real shift numbers in. See the full-loss number next to the reported one.
OEE = Availability × Performance × Quality. Nothing exotic. The gap comes from what usually gets left out.
Reported is a shortcut some shops use: (planned − logged downtime) / planned × quality.
It ignores micro-stops and never measures speed, so it flatters.
Full-loss counts the micro-stops as downtime and derives real speed from output:
Availability = (planned − logged − micro-stops) / planned,
Performance = (ideal cycle × total parts) / run time,
Quality = good / total.
What is an estimate: your micro-stop minutes and ideal cycle time. Without machine data, both are a guess. That guess is exactly the point — a real backbone measures them instead of estimating. Not sure of a number? Micro-stops show in the control's state history, if it keeps one, ideal cycle in the machine spec or the CAM post, quality in your inspection record.
We read shop-floor data for a living. A Factory Data Readiness Call is 30 minutes about your machines — where the number comes from, not a pitch. If it is not useful, you have lost half an hour.
Book a readiness call →Next tool: Readiness self-assessment — is your data ready to fix this?
These tools improve when practitioners push back. Tell us what you would change — a definition, an input, a default.
Send feedback