# Takt time

Demand-based production cadence, consistent available-time definitions, cycle-time comparisons and bounded line-balancing calculations.

Reviewed: 2026-11-02

Tags: production, planning, takt, capacity

In the worked shift, **480 min** minus **30** for a meal, **15** for a meeting and **15** for other planned stops leaves **420 min**. Demand of **280 accepted units** gives **90 seconds per unit** takt. Operator assignments of **110, 80 and 50 s** total **240 s**, yet the first exceeds takt. A hypothetical **25 s** transfer from the first to third gives **85, 80 and 75 s**. This static balance assumes no losses or waiting and tasks that can move as modeled.

Lean Enterprise Institute defines takt from available time and customer demand in the same period. [Lean Enterprise Institute, takt time](https://www.lean.org/lexicon-terms/takt-time/).

`takt time = available production time / required customer units`.

State the product family, output unit and calendar. Check actual cycle times and losses separately.

## Separate demand pace from observed performance

| Quantity | Meaning |
| --- | --- |
| Takt time | Available production time per required customer unit |
| Machine cycle time | Time for the machine’s operations on a piece |
| Operator cycle time | Time for an operator to repeat the assigned work elements |
| Production lead time | Elapsed passage through the production process, including waiting |

LEI distinguishes measured cycle time, machine and operator cycles, and production lead time. Its effective-machine-cycle definition also includes loading/unloading and an allocated share of changeovers. A cutting cycle displayed by a controller therefore need not represent the rate of completed, accepted output. [Lean Enterprise Institute, cycle time](https://www.lean.org/lexicon-terms/cycle-time/).

A long product lead time can coexist with a short interval between successive completions when several pieces are in process. Do not compare an entire job’s elapsed lead time directly with per-piece takt and conclude that one machine is too slow. First establish what each timing measures.

## State what “available” includes

Build the available-time figure from the calendar. Identify breaks, meetings, planned shutdowns, changeovers and any staffed relief arrangement explicitly. Minitab Engage’s documentation, for example, calculates production time from shift time less lunch breaks, meetings and similar exclusions. That is a documented calculation convention, not proof that every plant or software package uses the same boundary. [Minitab Engage, value-stream calculations](https://support.minitab.com/en-us/engage/help-and-how-to/tools/maps/map-and-timeline-calculations/).

Maintain the distinction between planned time available to meet demand and the subsequent losses during that time. If a loss has already been removed from available time, do not apply it again as a capacity factor. If it remains in available time, explain how the capacity plan will absorb it. Keep a customer-demand takt separate from any faster internal operating target chosen to accommodate losses.

For a particular example of terminology, Vorne’s OEE guidance distinguishes traditional demand-based takt from an “operable takt” tied to sustained process capability. Its planned-run-time convention excludes planned stops and nonscheduled time while recognizing losses during planned run time. Do not exchange these metrics without translating their definitions. [Vorne, calculating takt time](https://www.oee.com/takt-time/how-to-calculate/).

Zero demand gives no finite time-per-required-unit result. It is not an instruction to produce at an infinite interval while leaving an old schedule active. Resolve the demand and production plan explicitly. Similarly, missing demand or a missing calendar is an input problem, not evidence that a default takt is valid.

## Original example: calculate and balance a shift

Assume an invented eight-hour shift for one product family. The line stops for a 30-minute meal, a 15-minute meeting and 15 minutes of other planned nonproduction time. There is no relief production during those intervals. Customer demand is **280 accepted units per shift**.

`available time = 480 − 30 − 15 − 15 = 420 minutes = 25,200 seconds`.

`takt = 25,200 / 280 = 90 seconds per accepted unit`.

Assume each unit requires **240 seconds of manual work**, divided among three operators as 110, 80 and 50 seconds. There is no walking, blocking, machine waiting, scrap or rework in this initial ideal comparison; each operator has a dedicated station and all required material is ready.

Total work divided by takt is `240/90 = 2.667 operator equivalents`. Rounding up gives **three** as an arithmetic lower bound, not a staffing guarantee. The first operator’s 110-second assignment exceeds takt even though aggregate labor time appears sufficient. In a steady repeating line under these assumptions, its rate limits the line to `25,200/110 = 229.09 units per shift` before whole-unit and startup effects.

Suppose a technically feasible, approved change transfers a 25-second work element from the first operator to the third. The assignments become **85, 80 and 75 seconds**, still totaling 240. All are below the 90-second takt. At the stated demand, total manual work is `280 × 240 = 67,200 seconds`, compared with `3 × 25,200 = 75,600 operator-seconds` available. The difference is **8,400 operator-seconds**, distributed as 5, 10 and 15 seconds per takt interval across the three operators.

This reconciliation does not establish allowances for fatigue, safe access, relief, handling or variation. The transfer is an explicit hypothetical assumption; tasks may be indivisible or require different qualifications, tooling or sequence. LEI’s operator-balance-chart method compares individual work elements with takt and supports redistributing them rather than relying solely on the total. [Lean Enterprise Institute, operator balance chart](https://www.lean.org/lexicon-terms/operator-balance-chart/).

## A static balance does not prove delivery capacity

Continue the example with a separate forecast: the balanced line loses **10% of the 420 planned minutes** to interruptions and accepts **98% of completed starts**, with no rework. Its slowest 85-second station then permits a simplified expected accepted output of:

`25,200 × 0.90 / 85 × 0.98 = 261.49 units per shift`.

Even with all nominal station assignments below takt, estimated long-run output falls below 280. Takt stays at 90 seconds for the stated planned time and demand.

The example excludes additional startup and end-of-shift effects, shared operators, transfer delays and correlated downtime. Use representative observations or a suitable scheduling/simulation model when these effects matter. A stopwatch average alone is not a service guarantee.

## Product mix and review cadence

When products consume different work content, an aggregate unit count can conceal overload. For example, 200 simple units and 80 complex units are not necessarily interchangeable with 280 simple units. Retain the expected mix and routing times, then check each resource’s required work against its available time. A common family takt is useful only alongside that mix and a workable sequence.

Greg Lane’s authored high-mix/low-volume presentation describes difficulties in establishing takt for build-to-order work and emphasizes quoted times and real-time bottleneck management. Treat this as application guidance for that environment, rather than assuming a repetitive single-product line model fits every job shop. [Greg Lane, High Mix–Low Volume Lean, hosted by LEI](https://www.lean.org/wp-content/uploads/2021/06/82.pdf).

Review takt when the approved demand, product boundary or available calendar changes. Keep the basis stable enough that planned-versus-actual output is interpretable. A late shift should trigger diagnosis of losses and feasibility; repeatedly recalculating takt from achieved output would erase the gap the metric is meant to reveal.

For handoff, retain demand source/date, required good units, time reconciliation, takt, observed cycle definitions, work-element balance, loss assumptions and unresolved constraints. Distinguish the calculated pace from a capacity assessment, staffing decision and production commitment.

## Source versions

Reviewed **2026-11-02 PST**. No separately verified publication date is assigned to Lane’s PDF or the cited LEI lexicon, Minitab and Vorne pages. The PDF’s hosting path is not treated as a 2021 issue date. Vendor conventions are named where used. Numerical examples and operating recommendations are original and bounded to their stated assumptions.