What is the difference between theoretical and actual cycle time?

Theoretical (ideal) cycle time is the fastest a machine can make one good unit at full rated speed with no losses, a fixed standard. Actual cycle time is the real average, run time divided by total count, which absorbs every speed loss and minor stop during the run. Actual is always equal to or longer than theoretical, never shorter.

How do you calculate actual cycle time?

Divide run time by total count over a window when the machine was up and running one product. If it ran 60 minutes and made 1,440 parts, actual cycle time is 2.5 seconds per part. You only need two honest numbers, run time and count from a single discharge counter, not a sensor on every axis.

Why is actual cycle time always longer than theoretical?

Because it silently contains losses that theoretical excludes by definition. Reduced speed makes each part genuinely take longer, and minor stops spread the same count across more run time. Both are Performance losses, so actual cycle time can flag that a problem exists but cannot, on its own, tell you whether speed or stops is the cause.

How do you set theoretical cycle time correctly?

Anchor it to a demonstrated best sustained rate the line has actually held, revisit it when the process changes, and treat the manufacturer's nameplate as an unverified claim. The worst choice is 'what we usually run,' which bakes today's losses into the standard and hides them. A too-slow theoretical number is the most common way OEE flatters itself.

Theoretical vs Actual Cycle Time

Q: Is the ratio of theoretical to actual cycle time the same as the Performance factor?

Yes. The OEE Performance factor is ideal cycle time times total count divided by run time, which rearranges exactly to theoretical cycle time over actual cycle time. So a Performance of 80 percent means your actual cycle time is 25 percent longer than theoretical, one divided by 0.8.

Q: What does the gap between theoretical and actual cycle time tell you?

That Performance loss exists and roughly how large it is, but not yet why. A wide steady gap usually means a genuine speed problem; a gap that appears in bursts usually means minor stops; a gap worst right after every start points at startup losses. Reading the shape of the gap over time is the fastest triage available.

One ratio quietly decides a third of your OEE, and most plants never look at it directly. Here is the number behind the Performance factor.

Theoretical cycle time is the fastest a machine can make one good unit at its full rated speed with no losses; actual cycle time is the real average, run time divided by the count it produced. Divide theoretical by actual and you have the OEE Performance factor. The two numbers are the same measurement, ideal versus real.

This one ratio quietly decides a third of your OEE, and most plants never look at it directly. They compute Performance from a formula and move on, never noticing that the whole factor is just theoretical cycle time over actual cycle time. Once you see it that way, the Performance number stops being abstract: it is the ratio of the clock the machine should keep to the clock it actually keeps. This post defines both, derives the ratio, and shows how to measure actual cycle time without wiring up every axis on the line.

What is theoretical cycle time?

Theoretical cycle time, also called ideal cycle time, is the minimum time to produce one unit when the machine runs at full rated speed with zero stops and zero slow-downs. It is the denominator of your rate: if the machine is capable of one good part every 2.0 seconds under perfect conditions, that 2.0 seconds is the theoretical cycle time, and it is the yardstick every Performance number is measured against.

Where you get that number decides whether your whole OEE is honest. Three common sources, in rough order of trustworthiness: the demonstrated best sustained rate the line has actually held for a real run, the equipment manufacturer's nameplate rating, and, worst, "what we usually run," which bakes today's losses into the standard and makes them invisible. Anchor theoretical cycle time to a demonstrated best, revisit it when the process changes, and treat nameplate as an unverified claim. This is the same discipline that governs cycle time generally, sharpened for the number that sets your rate.

What is actual cycle time?

Actual cycle time is the real average time per unit over a window when the machine was running: run time divided by total count. If the machine ran for 60 minutes and produced 1,440 parts, actual cycle time is 2.5 seconds per part, whatever the theoretical says. It is always equal to or longer than theoretical, never shorter, because it silently absorbs every speed loss and every micro-stop that happened during the run.

That is the key property: actual cycle time is a container for two of the six big losses at once. Reduced speed stretches it because each part genuinely takes longer; minor stops stretch it because brief halts spread the same count across more run time. Actual cycle time does not tell you which of the two is hurting you, only that the gap exists. Splitting the gap is the job of speed loss diagnosis but you cannot start until you have measured the gap itself.

Same run time, two spacings. The theoretical row fits ten parts, the actual row eight. The ratio of the two cycle times, 80 percent, is the Performance factor, and the missing two parts are the loss.

Why is their ratio the Performance factor?

Because Performance is defined as ideal cycle time times total count, divided by run time, and that rearranges to theoretical cycle time over actual cycle time. Actual cycle time is run time divided by count, so dividing theoretical by actual gives theoretical multiplied by count over run time: the exact Performance formula. The algebra is short, and it is worth doing once because it turns an opaque factor into something you can reason about at the machine.

The practical payoff of seeing it as a ratio is intuition. A Performance of 80 percent means your actual cycle time is 25 percent longer than it should be, one over 0.8. If someone tells you the line runs "a little slow," you can ask for the two cycle times and know exactly how little. And because OEE multiplies its three factors, this ratio flows straight through: recover half the gap between actual and theoretical and you add points to the whole OEE score, not just to Performance.

The ratio also settles arguments that opinions never will. "The line feels slow today" becomes: theoretical is 2.0 seconds, actual came in at 2.6, Performance is 77 percent, and we are losing roughly 0.6 seconds on every part. That is a claim someone can act on and check tomorrow. Compare it to the wider family of rate measures, flow rate and takt answer different questions, and the theoretical-to-actual ratio is the one that speaks specifically to how hard the machine itself is being pushed versus what it can do.

The same factor written three ways. Seeing Performance as a plain ratio of two cycle times makes it something you can check by hand at the machine.

How do you measure actual cycle time without a fleet of sensors?

You need exactly two honest numbers, run time and count, not a sensor per axis. Work it in this order:

Pick a clean window. Choose a span where the machine was up and running one product. Exclude planned downtime and changeovers; you are measuring how fast it ran while it ran, not availability.
Record run time. The actual minutes the machine was running in that window. A run/idle signal, a PLC bit, or even a disciplined manual log will do, precision matters more than fancy instrumentation.
Count the units. Total pieces produced in the window from a single counter on the discharge. Keep good and total both if you can, so the same data also feeds your Quality factor.
Divide for actual cycle time. Run time divided by total count. That is your real average time per part, minor stops and speed losses and all.
Compare to theoretical. Divide theoretical by actual for the Performance factor, or subtract to see the seconds lost per part. Either way, the gap is now a number instead of a feeling.
Watch it over time, not once. A single snapshot is an anecdote. Trend actual cycle time run over run and the gap's pattern, steady, or spiking at startups, or drifting as tooling wears, points straight at the cause.

The definitions this rests on:

The OEE Performance factor is formally ideal cycle time multiplied by total count, divided by run time, algebraically identical to ideal cycle time over actual cycle time. The formula and the treatment of actual cycle time as the container for speed loss are documented in the OEE calculation reference.
The gap between the two cycle times is filled by exactly two of the six big losses, reduced speed and minor stops, both classed as Performance losses in the six big losses reference. That is why actual cycle time can flag a Performance problem but cannot, by itself, tell you which of the two to chase.

How do the two cycle times compare at a glance?

They differ in what they measure, where they come from, and what moves them. The table lays them side by side.

Aspect	Theoretical cycle time	Actual cycle time
What it is	Fastest possible time per unit	Real average time per unit
How you get it	Demonstrated best rate or nameplate	Run time ÷ total count
Changes when	The process or spec changes	Every run, with losses
Contains losses?	No, losses excluded by definition	Yes, speed loss and minor stops
Role in OEE	The rate standard (denominator)	The measured reality (numerator side)

Theoretical is the fixed standard; actual is the moving measurement. The ratio between them is the Performance factor, and the distance between them is your opportunity.

What does the gap tell you?

That there is Performance loss, and roughly how much, but not yet why. A wide, steady gap usually means a genuine speed problem: worn tooling, a conservative derate, a downstream starve. A gap that appears in bursts usually means minor stops. A gap that is worst right after every start points at startup losses bleeding into the ramp. Reading the shape of the gap over time is the fastest triage you have, and it costs nothing beyond the counts you are already able to collect.

The one thing that will wreck the whole analysis is a bad theoretical number. Set it to yesterday's comfortable rate and the gap closes on paper while the loss stays real, the single most common way OEE flatters itself, and one of the common OEE mistakes worth auditing for. Anchor theoretical to a demonstrated best, measure actual from clean counts and run time, and trend the ratio as one of your core manufacturing KPIs. Plants that stream counts and run/idle straight off the line the way Harmony does (see the platform) get actual cycle time continuously instead of from a once-a-shift stopwatch, and you can turn the recovered gap into a number with the OEE calculator. The two cycle times were always there. Measuring both is what turns "runs a little slow" into a plan.