Takt time
The rhythm the customer buys at: available working time divided by demand. Not a measure of how fast the line runs — a target it should match.
Formula
In practice
Takt is a statement about the customer, not about the plant. It is computed from unreduced available time — what the shift pattern offers before any allowance for how well the equipment runs — because capability belongs in capacity, and folding it in here would produce a rhythm the plant is already comfortable with and call it demand.
Weekly and monthly quantities are spread over operating days rather than calendar days. A customer ordering 1,500 a week from a five-day plant needs 300 a day, not 214, and the difference is the whole of a capacity argument. Breaks, planned maintenance and meetings come out of available time; breakdowns and micro-stops do not, because those are counted in OEE availability and charging them twice would understate the day.
How this is calculated
demand per working day = quantity / period, spread over operating days
takt = net available / demand per working day
pitch = takt × pack quantity
Weekly and monthly demand is spread across operating days, not calendar days: a customer ordering 1,500 a week from a five-day plant needs 300 a day, not 214.
Takt uses unreduced available time. It expresses what the customer requires, not what the plant can manage — capability enters through capacity, below.
Common questions
Is takt time the same as cycle time?
No, and comparing them is the point. Takt comes from demand and the calendar; cycle time is derived from the work elements at a station. A process whose cycle time exceeds its takt cannot keep up, which is the finding the Acme Stamping example exists to produce.
Does every process on the map share one takt?
Only where nothing is scrapped or reworked. Yield loss means upstream processes must start more than the customer receives, so each process gets a yield-adjusted takt of its own. Testing a stream against a single figure stops being valid the moment yield drops below 100%.
This definition is the one the product itself shows, read from its interface catalogue at build time.