How to Calculate OEE: Formula, Worked Example and Shift Template

Calculate OEE by multiplying availability, performance, and quality, using the same machine, product, and time period for all three. Record planned production time, stop time, ideal cycle time, total output, and first-pass good output. The example below shows every input, the calculation, and how to turn the result into a daily improvement decision.

Reviewed: 8 September 2026. The worked example is illustrative; it is not an AICAN customer result or an industry benchmark.

OEE formula and the five inputs you need

Overall Equipment Effectiveness measures productive output against the output a process could produce during its planned production time. The standard calculation is OEE = Availability × Performance × Quality. Vorne’s OEE calculation reference documents the formulas; its OEE factors guide explains the loss categories.

Input	What to record	Owner to agree before starting
Planned production time	The time the selected process was intended to produce	Production planner and shift supervisor
Stop time	Stops inside that planned production window	Operator, with supervisor review
Ideal cycle time	The validated fastest cycle for the specific product, in a consistent unit	Process engineering
Total count	All units produced during the same window	Counter or production record
Good count	Units meeting quality requirements the first time	Quality team

Before comparing shifts, agree where counting starts and ends. For example, counting parts after inspection on one shift and before inspection on another creates a misleading comparison. Write the measurement boundary on the shift sheet, including the machine or line, product, shift start, shift end, and person responsible for review.

A complete worked example for one machine

Imagine a component-making machine scheduled for a 480-minute shift. The factory has explicitly planned 30 minutes without production, leaving 450 planned production minutes. During that production window, the machine stops for 60 minutes. It makes 340 components, of which 323 pass inspection the first time. The validated ideal cycle is one minute per component.

Calculation	Working	Result
Planned production time	480 − 30 minutes	450 minutes
Run time	450 − 60 minutes	390 minutes
Availability	390 ÷ 450	86.67%
Performance	(1 minute × 340 components) ÷ 390 minutes	87.18%
Quality	323 ÷ 340	95.00%
OEE	(390 ÷ 450) × (340 ÷ 390) × (323 ÷ 340)	71.78%

The arithmetic has a useful cross-check: 323 good components × one ideal minute ÷ 450 planned minutes also equals 71.78%. Calculate with the unrounded inputs, then round the final display. Mixing rounded percentages can create small differences that waste time during reviews.

Here, 450 planned minutes produced 323 ideal minutes of good output. The 127-minute difference consists of 60 minutes of stops, 50 minutes of speed loss, and 17 ideal minutes associated with output that did not pass first time. This breakdown comes from the example’s inputs, not from a general assumption about factories.

Copy this layout into a shift spreadsheet

Put the following labels in column A and values or formulas in column B. Use one sheet per consistently defined reporting period, or add separate rows for each machine and product. These formulas work in Excel and Google Sheets with numeric inputs.

Cell	Label	Value or formula
B2	Shift minutes	480
B3	Time not scheduled for production	30
B4	Stop minutes during planned production	60
B5	Ideal minutes per component	1
B6	Total components	340
B7	First-pass good components	323
B8	Planned production minutes	=B2-B3
B9	Run minutes	=B8-B4
B10	Availability	=IF(B8>0,B9/B8,"")
B11	Performance	=IF(B9>0,B5*B6/B9,"")
B12	Quality	=IF(B6>0,B7/B6,"")
B13	OEE	=IF(AND(B8>0,B9>0,B6>0),B10*B11*B12,"")

Format B10:B13 as percentages. Blank results indicate an undefined calculation requiring review; they should not silently become a perfect score or a zero. Add validation so good count cannot exceed total count, stop time cannot exceed planned time, and cycle time must be positive. Keep the raw records alongside the calculation so a supervisor can explain an unusual result.

Use the result to choose the next action

Start with the largest documented loss that the team can investigate. In this example, stops account for 60 minutes. A useful review would separate those stops by reason: breakdown, changeover, missing material, waiting for instructions, or another agreed code. The largest reason becomes a candidate for investigation, with an owner and a check date.

Suppose the stop log shows repeated waits for one bought-out component. The next action could be to reconcile that item’s physical stock, open purchase orders, and expected consumption before the next shift. That connects the OEE discussion to inventory visibility and the production plan. It does not require an AI prediction to make the first useful improvement.

After the action, compare equivalent periods and preserve the definitions. Record whether the targeted stop reduced, whether another loss increased, and whether delivery or quality improved. A higher percentage alone does not prove that the business result improved.

Data checks before you trust the dashboard

• Use matching units. Convert seconds to minutes before combining cycle time with run time.
• Separate products with different cycles. Do not reuse a fast product’s cycle time for every part number.
• Investigate impossible values. Performance above 100% is a reason to inspect cycle standards, counters, and timing boundaries.
• Keep downtime definitions stable. Do not improve the score by moving inconvenient stops outside the planned window.
• Record corrections. A counter reset, missing quality entry, or duplicate production booking should have a visible explanation.
• Review input completeness. Mark a shift as incomplete when a necessary count or time record is unavailable.

Put Optiwise to work in your factory

Connect machine and process data in Optiwise, using IoT inputs alongside operator or supervisor entries. Capture time, output and quality results against production work. For the OEE calculation above, define the planned production time, ideal cycle time and good-count rules for the machine and product being measured.

You can discuss your factory’s measurement setup with AICAN. Confirm the equipment connections and workflow scope during that discussion; this guide does not promise that every machine exposes every required signal.

Frequently asked questions

Can I calculate OEE without IoT sensors?

Yes. A consistent shift log containing planned time, stop time, cycle time, total count, and first-pass good count is enough to calculate it. Automatic collection can improve timeliness, but the definitions and review process still need to be agreed.

Should I average machine OEE percentages?

Do not treat a simple average as a reliable plant result. Machines with different planned times and product cycles need a consistent aggregation method and a clearly defined purpose. Keep individual results visible before introducing a combined number.

Is an OEE target of 85% right for every factory?

No single percentage is an appropriate target for every process. Establish a reliable baseline, understand the product mix and constraints, and select improvements that help throughput, quality, delivery, and cost.

What should I do when a shift has no planned production?

Record that the period had no planned production and leave the OEE calculation undefined. Do not divide by zero or count the period as perfect performance.