Blog / OEE
OEE · Guide

OEE Calculation for a Machine Shop, by Example

OL By OEE Lab Editorial |Updated 2 November 2026 |5 min read
Short answer

Shop OEE = total good ideal hours / total scheduled hours, added across the machines you include. Good ideal hours are good parts x proven time per part, summed from your job records. Do not average each machine's percentage, because a machine scheduled for 20 hours would count as much as one scheduled for 80.

From one machine to the whole shop

If you own a dozen machines, sooner or later someone asks for one number. This page shows how to build it from the paperwork most shops already have.

For one machine's rules, read OEE for CNC machines. For several jobs on one machine, read OEE for job shops.

Step 1: choose which machines count

Leave out machines that only run when someone needs them, such as a deburr bench or a small saw. Their waiting time would drown the number in spare capacity.

Include the machines jobs queue for. A manual mill belongs in if it is booked like the others, but only for the hours it was actually scheduled.

Step 2: set scheduled hours per machine

Scheduled hours are the hours you planned to run that machine, minus planned breaks. A machine with no job planned for the second shift is not scheduled for the second shift.

That rule stops an unbooked machine from pulling the shop down. Use the same rule every week, and see OEE with no orders for the slow weeks.

Step 3: turn job records into ideal hours

A job traveler or ERP labor ticket usually holds the part, the machine and the quantities. Add the proven time per part, and you have what you need.

JobGood partsProven min per partGood ideal hours
4417 bracket406.040 x 6 / 60 = 4.0
4422 shaft1210.012 x 10 / 60 = 2.0
4430 housing822.58 x 22.5 / 60 = 3.0
Hours are the only thing you can add across different parts. Hypothetical jobs.

Use the time you have proven, not the time you quoted. Quotes carry margin and allowances, so a part quoted at 9 minutes and proven at 7.5 would inflate this job's ideal hours by 20%.

For a new part with no proven time, use the best run so far and mark it provisional. Update it when the part repeats, using how to find the ideal run rate.

Step 4: add the hours, then divide

Example: a five-machine shop in one week. The fourth machine runs one shift only, and the manual mill is booked 20 hours.

MachineSched hStops hRun hIdeal h madeGood ideal hOEE
VMC 1802060545265.0%
VMC 2802456484657.5%
Lathe801466585670.0%
HMC40634313075.0%
Manual mill2061410945.0%
Shop3007023020119364.3%
Stops include setups. Hypothetical numbers.

Availability is 230 / 300 = 76.7%, performance is 201 / 230 = 87.4% and quality is 193 / 201 = 96.0%. Multiplied, those give 64.3%, the same as 193 / 300.

The plain average of the five percentages is 62.5%. The manual mill's 45% counts as much as a full VMC there, as how to average OEE across machines explains.

Good ideal hours193 h, 64.3%
Stops, including setups70 h, 23.3%
Slow running and short stops29 h, 9.7%
Scrap8 h, 2.7%
Where the shop's 300 scheduled hours went: 193 + 70 + 29 + 8 = 300.

The ranking is the useful part. Setups and other stops take 70 hours, so shortening setups is worth more here than a faster toolpath on any one part.

What is different in a machine shop

  • One operator, two machines. One of them often waits. That waiting is a stop on the machine, so log it, and don't treat it as a personal failing.
  • Prove-out parts. First parts that get scrapped count against quality, and in short runs they hit hard.
  • In-cycle tool changes. They belong inside the proven time. Only breakage and unplanned changes are stops.
  • Manual machines. Times vary by operator and setup, so use wider tolerance and don't read one week closely.
  • Lights-out hours. If the shop schedules them, they count. Good unattended hours lift the score, and an alarm at 2 am shows up as a long stop.

Check your data before you trust it

Machine monitors can make a shop look better than it is. In a study of 884 machines in 23 Swedish companies, Hedman and co-authors (2016) found the companies had not logged design speeds.

Most machines showed 100% for performance and quality, which was the system's default value.

The same study put machines in mechanical workshops among its groups, with 364 machines from 9 companies. The lesson for your shop is simple: a performance figure of exactly 100% week after week is a missing input, not a win.

Spindle-on time is another trap. It tells you the machine was cutting, not that it was cutting at the proven rate or making good parts, as utilization vs OEE shows.

Reading the number

Do not expect a job shop to match a dedicated line, since setups put a ceiling on the score. How to set OEE targets shows how to build a target from your own baseline.

Look at the machine with the longest queue first, using bottleneck analysis. A gain on a machine with spare time ships nothing extra.

Common mistakes

  • Quoted hours as the ideal. They hide speed loss and can push performance over 100%.
  • Counting paid hours as scheduled hours. A machine with no job should not count as scheduled.
  • Mixing rework in. A part remade on a second traveler can count twice unless you mark it.
  • Ranking machines by OEE alone. A high score on a light-duty lathe does not make it more important than the VMC everything waits for.
See it live in Fabrico

A shop usually mixes controls of different ages. Fabrico reads stops, speed and counts live from PLCs, or from external IoT sensors on older machines, and gives OEE per line and shift. See how Fabrico connects to older machines.

What people ask next

How do you calculate OEE for a machine shop?

Add good parts x proven time per part across every job and machine, then divide by total scheduled hours. The plant OEE calculator does the weighted roll-up for you.

Should every machine be included?

No. Include the machines jobs queue for, and leave out those that only run on demand. Start with the constraint, using bottleneck analysis to find it.

What if I do not know the ideal cycle time for a part?

Use the best time you have proven on that part, not the quote, and mark it as provisional. How to find the ideal run rate gives the steps.

Does setup time count against OEE in a machine shop?

Yes. A setup is a stop inside scheduled time even when it was planned. How to count changeovers in OEE explains why.

Can I track shop OEE in a spreadsheet?

Yes, one row per job with scheduled hours kept per machine works well to start. The OEE Excel formula guide has the setup.

See OEE per line and shift on your own machines

Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.

Book a Fabrico demo

← More OEE Lab guides · Open the plant OEE calculator →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES