Overall Equipment Effectiveness: Calculate and Use OEE

Home » Resources » Overall Equipment Effectiveness: Calculate and Use OEE

Estimated reading time: 10 minutes

Overall equipment effectiveness (OEE) measures how much of your planned production time becomes good output at the process’s ideal speed. It combines three factors: availability, performance and quality. A machine can run for most of a shift and still have a low OEE because it runs slowly or produces parts that need correction.

That makes OEE useful when the factory feels busy but deliveries keep slipping. It separates lost running time from lost speed and lost first-pass quality, giving you a better starting point than simply asking the team to produce more.

The next step is to act on the loss. SkyPlanner brings an OEE KPI into the production overview alongside capacity and load, with AI scheduling in the same environment. You can investigate the result, account for the confirmed capacity change and revise the connected production plan without rebuilding each machine queue by hand.

How to calculate OEE

The formula is:

OEE = Availability × Performance × Quality

Use decimal fractions when multiplying, then express the result as a percentage. Do not add the three scores or take their average: each factor reduces what remains after the previous loss.

Factor	Calculation	Question it answers
Availability	Run time ÷ Planned production time	Did the process run when we intended it to?
Performance	Ideal cycle time × Total count ÷ Run time	While running, did it produce at its ideal speed?
Quality	Good first-pass count ÷ Total count	How much output met requirements without rework?

The time and quantity records must describe the same equipment, production window and output boundary. A machining result is not automatically the result for the entire route through finishing and assembly.

Availability: time lost to stops

First define planned production time: the time during which you intended the equipment to produce. Exclude periods with no intention of production, such as an unstaffed shift or a break when the process is scheduled to stop.

Then subtract stops within that production window to find run time. Breakdowns reduce availability, but so do planned changeovers within the window. Knowing about a setup in advance does not make its production time disappear.

Agree a consistent rule for which stops are recorded as downtime. Longer interruptions normally belong in availability; brief stops that are not separately logged remain in the performance loss. Count each interruption once, not in both factors.

Performance: output lost while running

Ideal cycle time is the fastest achievable time per unit for that product under optimal conditions. It is not the quoted lead time or a scheduling allowance padded for stops and setups.

Multiply ideal cycle time by the total quantity produced, including defective units, and divide by run time. Use matching time units throughout. Slow cycles and small stops reduce the result even when the equipment appears to be running normally.

A performance result above 100% is a reason to check the ideal cycle time, quantities and time records. Raising the ideal cycle time makes the score look better without making the process faster.

Quality: output that was right the first time

Good count means output that met the agreed requirements without rework. A rejected part that later passes after correction is recovered output, not a first-pass success for the original production result.

Keep total produced, first-pass accepted, scrap and rework quantities distinguishable. Otherwise, a healthy final accepted total can hide the extra machine time needed to achieve it. First pass yield explores that quality question in more detail; OEE combines it with the time and speed losses.

OEE calculation example: where did the shift go?

Imagine one machine producing the same component throughout an eight-hour shift. There is no production planned during a 30-minute break. The remaining records are:

Shift record	Value
Shift length	480 minutes
Scheduled break without production	30 minutes
Planned production time	450 minutes
Recorded stops, including changeovers	90 minutes
Run time	360 minutes
Ideal cycle time	0.5 minutes per part
Total parts produced	600
Parts accepted without rework	570

Availability is 360 ÷ 450 = 80%.

Performance is (0.5 × 600) ÷ 360 = 83.33%.

Quality is 570 ÷ 600 = 95%.

Multiplying the unrounded factors gives 63.33% OEE. You can check it with the shorter, equivalent formula:

OEE = Good count × Ideal cycle time ÷ Planned production time

For this shift, that is 570 × 0.5 ÷ 450 = 63.33%.

The 570 good parts represent 285 minutes of production at the ideal cycle time. The other 165 minutes within planned production time were lost to stops, reduced speed and defects. They are not all downtime.

Compare the losses in minutes, not just percentages

The three factor percentages have different denominators. Subtracting each from 100% does not give three directly comparable shares of the shift. Convert them to the same ideal-production-time basis:

Loss	Calculation	Lost minutes
Availability loss	450 planned minutes − 360 run minutes	90
Performance loss	360 run minutes − (600 parts × 0.5 minutes)	60
Quality loss	30 parts not accepted first time × 0.5 minutes	15
Total loss	90 + 60 + 15	165

The 15 quality-loss minutes are an ideal-time equivalent, not a measurement of the rework required. Correcting those parts could take much longer or use a different resource.

In this example, recorded stops are the largest time loss. Start by examining their reasons and duration. If most of the 90 minutes came from a recurring changeover problem, setup improvement deserves attention. If a single breakdown caused the loss and the machine is still unavailable, the immediate planning decision is different: protect affected deliveries while maintenance restores it.

Turn the OEE factors into an improvement decision

Availability, performance and quality tell you what kind of loss occurred. They do not establish its root cause.

A useful investigation connects the factor to an observed event. Equipment failures and setup or adjustment losses affect availability. Minor stops and reduced speed affect performance. Startup rejects and defects during production affect quality. These are commonly grouped as the six big losses.

Use the equipment records and the people doing the work to narrow the explanation. Repeated short interruptions might come from feeding problems, while a slower cycle could reflect worn tooling or a difficult product variant. The same percentage can conceal very different problems.

Choose an action by its production consequence as well as its size. A recurring loss on the resource limiting a due order may deserve attention before a larger loss on a machine with ample spare capacity. More output is useful when downstream operations and customer demand can use it; producing unwanted parts just to improve a local score creates inventory, not better delivery performance.

Give the chosen action an owner and a check on comparable production. Maintenance, a better changeover method or a process correction addresses the loss itself. Scheduling addresses the work and capacity available while that improvement is being made. You need both when an equipment problem threatens an order already promised.

OEE vs. utilization: busy is not the same as effective

Utilization asks how much of a defined resource capacity is used. Its denominator may be available shift hours or another stated time base. OEE asks how effectively the intended production time became good output at ideal speed.

A heavily loaded machine can therefore have poor OEE. It may spend most of its running time producing slowly or making defects. Equally, a machine can achieve strong OEE during its scheduled production window while standing unused outside that window because there is no demand.

OEE alone does not measure use of every hour in the calendar. Total effective equipment performance (TEEP) extends the picture to calendar time by including the proportion scheduled for production. For day-to-day decisions, keep the questions clear: is there enough demand and scheduled work, and how much productive output are we getting from the time we intend to run?

Is 85% OEE a good target?

The often-quoted 85% figure is a familiar discrete-manufacturing benchmark, not a universal pass mark. A high-mix machine with frequent necessary changeovers is not directly comparable with a dedicated line running one product.

Start with a trustworthy baseline for your own process. Keep the production-time boundary, ideal cycle times and acceptance rules consistent, then target a confirmed loss you can reduce. A better number caused by excluding troublesome time or counting reworked parts as first-pass good output is not an operational improvement.

When products have different ideal cycle times, calculate their productive-time contributions using the correct rate for each product. Aggregate the underlying time and output data rather than taking a simple average of unrelated shift or machine OEE percentages.

Use SkyPlanner to connect OEE with the next production plan

A shift review is more useful when the equipment result and the delivery commitments are not in separate worlds. SkyPlanner’s Home dashboard includes OEE alongside capacity, load and production-status KPIs. Job reports and planned-versus-realized reporting help you examine the work behind the overview, while Timer records work time and produced or faulty quantities.

Once you understand a loss, make its planning consequence specific. An unavailable machine needs an availability change. A recurring slower process needs a realistic operation duration. Approved rework needs its own resource time. A single composite percentage cannot tell you which of those changes to make.

For the example shift, suppose maintenance confirms that the machine also needs a two-hour repair tomorrow. Record that unavailable window as a workstation exception in SkyPlanner. The exception blocks scheduling during that time. Arcturus, SkyPlanner’s scheduling AI, can then recalculate the affected work against the configured capacity, operation dependencies and priorities; Gantt shows the resulting sequence and completion dates.

You can see what the lost window means for the next operations and other orders sharing the machine, rather than merely moving tomorrow’s first job and discovering the knock-on effects later. If another configured workstation is eligible, the schedule can account for its capacity too. You retain control of priorities and protected decisions; Arcturus handles the connected scheduling calculation.

Keep OEE’s ideal-speed reference separate from the realistic durations used to plan work. If the schedule already includes actual processing rates, setup time and known downtime, applying the whole OEE percentage again can count the same losses twice. Update the input that reflects the confirmed problem instead.

SkyPlanner integrates with any ERP or MES through its open REST API. That lets you connect existing order and production data with finite scheduling, reduce separate KPI administration and spend less time repairing machine queues when capacity changes.

What a better work sequence changed at Mamec

Mixing-technology and contract manufacturer Mamec previously prioritized thousands of interdependent jobs using paper work orders and its ERP’s rigid planning functions. The overall load was difficult to see, and production control carried the burden of coordinating the sequence.

With SkyPlanner integrated with Oscar ERP, Mamec moved to AI-powered finite scheduling and shared visibility of orders and load. The company reports less workstation downtime from an optimized work sequence and a clearly lighter planning workload.

That is a practical reason to connect equipment performance with scheduling: the next job and its dependencies become part of a maintained plan, rather than another set of local decisions to reconcile.

See what your equipment losses mean for delivery

Choose one machine whose lost time regularly disrupts promised work. Bring its production window, ideal cycle time, quantities and main stop reasons to a SkyPlanner demo, together with an affected order and its remaining operations.

Use that situation to explore the whole response: understand the OEE result, record a confirmed capacity change and see how the revised schedule affects deliveries. Book a SkyPlanner demo to discuss your production, or try SkyPlanner with demo data first.