What is OEE? A simple explanation for one machine

The three parts of OEE

OEE splits machine losses into three factors. Each factor is a percentage. You multiply them together.

Availability measures time. It compares the time the machine actually ran against the time you planned to run it.

Performance measures speed. It compares the parts you made against the parts the machine could have made at the ideal cycle time while it was running. Slow cycles eat performance. Small stops that are not recorded as availability losses eat it too.

Quality measures good output. It compares good parts against all parts made. Scrap eats quality. Parts that require rework are not first-pass good parts.

The formulas:

Two definitions matter. Planned production time is the time you scheduled the machine to produce. Define which breaks and other exclusions fall outside that time, then keep the rule consistent. Ideal cycle time is the shortest achievable time to produce that part under optimal operating conditions. It is not the average cycle time or the pace implied by a daily target.

Performance above 100% signals that the ideal cycle time or another input needs checking. A machine that is 90% available, runs at 90% of its ideal production rate and makes 90% good parts scores 0.9 × 0.9 × 0.9 = 72.9% OEE. Small losses stack.

Here is how to assemble the inputs for a separate OEE worksheet when a camera supplies the production count:

Availability = run time ÷ planned production time
Performance = (total count × ideal cycle time) ÷ run time
Quality = good count ÷ total count
OEE = availability × performance × quality
What it measures (Availability: Time lost to stops; Performance: Output lost to slow running; Quality: Output lost to defects)
Formula input (Availability: run time ÷ planned production time; Performance: (total count × ideal cycle time) ÷ run time; Quality: good count ÷ total count)
Source of the number (Availability: Your shift schedule and stop records; Performance: Camera count and the ideal cycle time for the part; Quality: Your inspection records, including scrap and rework)

The three factors describe different types of production loss.

One machine, one shift: a worked example

This section is an illustration. The numbers are invented to show the arithmetic. They are not a measured result from any factory. Sample views on the FactoryVision site also use example data.

Take one machine and one 8-hour shift. For this example, all 480 minutes are planned production time.

Cross-check from the other direction: 76 good parts × 5 minutes = 380 minutes of fully productive time. 380 ÷ 480 ≈ 79.2%. Same answer.

Availability is 12.5 percentage points below 100%; performance and quality are about 4.8 and 5.0 points below 100%. These gaps are not additive because OEE multiplies the factors. Under the 5-minute ideal-cycle assumption, 76 good parts represent 79.2% of the ideal output possible during the 480 planned minutes.

Planned production time: 480 minutes.
Downtime: 60 minutes, split between a tool change and waiting for material.
Run time: 480 − 60 = 420 minutes.
Availability: 420 ÷ 480 = 87.5%.
Ideal cycle time: 5 minutes per part.
Output possible in 420 minutes: 420 ÷ 5 = 84 parts.
Illustrative camera count: 80 parts.
Performance: 80 ÷ 84 ≈ 95.2%.
Rejected parts in this example: 4.
Good parts: 80 − 4 = 76.
Quality: 76 ÷ 80 = 95.0%.
OEE: (420 ÷ 480) × (80 ÷ 84) × (76 ÷ 80) ≈ 79.2%.

Worked example from the article; illustrative numbers, not a measured factory result.

Planned shift time splits into recorded run time and downtime in this example.

From a daily count to OEE

FactoryVision counts finished work with a camera on each machine. Each station gets one status light against the daily target: on track, at risk, or behind. The status is visible across the factory floor. Production is tracked by station and by shift, alongside the orders waiting on it. Counting requires no additional sensors or wiring into the machine.

For the OEE calculation described here, use a separate spreadsheet or paper worksheet. Gather:

Use the same machine, part and time period for all inputs. FactoryVision supplies the production count; your schedule, stop records and inspection records supply the other inputs for this worksheet.

FactoryVision’s models run on a small computer inside your factory, not in FactoryVision’s cloud.

Planned production time, from your shift schedule.
Run time, calculated from planned production time and your stop records.
Ideal cycle time for the part. Define it independently from the daily target.
Total count, supplied by the camera.
A reliable good-part count, or records of rejected parts, including those requiring rework.

The limits: what the camera cannot see

FactoryVision fits countable discrete parts. It is not an ERP, MES or SCADA system. Camera-based inspection has clear limits:

Visual QA checks parts while they are made. Its checks cover dimensions, appearance, completeness, matching against part references, process drift and cycle-time statistics. Each station uses the checks it needs.

OEE’s quality factor requires a good-versus-total count. In this article’s worksheet, use your inspection records to establish that count, including checks for defects the camera cannot see.

It cannot see defects inside material.
It cannot see the back of a part without a second camera.
It cannot measure finer than about 0.1 mm with reasonable optics.
Shiny or transparent parts need suitable lighting.
Continuous processes such as extrusion are outside its scope.

Frequently asked questions

What is OEE, in one sentence?

A percentage that combines availability, performance and quality to show how close production came to its ideal output during planned production time.

Can OEE be calculated with no sensors and only hand-entered numbers?

Yes. OEE is arithmetic using planned production time, run time, ideal cycle time, total count and good count. Record those inputs by hand in a separate worksheet. A camera supplies the production count.

Does a small workshop need OEE measurement?

Not always. As practical guidance, a daily count against a daily target can be a useful starting point. Use OEE when you need to separate time, speed and quality losses or compare similar shifts on the same machine.

Does the quality part of OEE come from a camera automatically?

For this worksheet, calculate quality from your good-part and total counts. FactoryVision’s Visual QA checks visible characteristics of parts. Your quality records must also account for defects outside the camera’s view, including internal defects.

How do we connect the result to the daily production count?

Use the shift count as total count in a separate OEE worksheet. Add planned production time, run time, ideal cycle time and good count for the same period. Keep the daily target separate from the ideal cycle time.

Which checks fall outside the camera’s scope?

Internal defects, the back of a part without a second camera, and measurements finer than about 0.1 mm with reasonable optics. Continuous processes such as extrusion are also outside FactoryVision’s scope.

How is setup carried out?

Mount and align the camera, position the printed reference marker, run ten parts and confirm the detected cycle. Show good parts and mark the features to check. A week of quiet observation precedes activation by your quality lead.

Where do the models run?

On a small computer inside your factory, not in FactoryVision’s cloud.

Book an introductory call

FactoryVision runs in a Turkish door and furniture factory today and is available to buy. It fits factories that make countable discrete parts, with 10 to 100 machines its best-fit range. The first step is an introductory call. After the call, the team gets to know your company so the setup fits your machines and processes. Book the introductory call

Request an intro call