Bottleneck Identification: Find the Constraint, Free the Line
The constraint sets the pace
Every line has a constraint — the one step whose capacity limits the whole line's output. Improve any other step and the line produces nothing more; improve the constraint and the whole line moves. That is the whole theory, and it is brutally practical: the constraint is where your improvement money belongs, and everywhere else is, for now, a distraction. Plants waste enormous effort speeding up non-constraints, then wonder why output did not move. Output moves when the constraint moves. Everything else is motion disguised as progress.
Four ways to spot it
You do not need software to find a bottleneck. Four floor signs will point at it. First, the queue: work piles up before one station and nowhere else. Second, starved and blocked: the station before the constraint is blocked because it cannot pass work forward, and the station after it is starved because nothing arrives. Third, overtime: one station consistently works late while the rest of the line goes home on time. Fourth, inventory: the biggest pile of work in process on the line sits directly in front of it. When two or more of these signs point at the same station, you have found your constraint. When they point at different stations, look again — the real one is the station the others are waiting on.
Don’t confuse busy with bottleneck
The busiest station is not always the bottleneck, and the most complained-about station is not always the constraint. A station can look frantic because it is poorly organized, badly supplied, or running a bad sequence — and fixing that chaos may help the station without moving the line's output at all. Conversely, a quiet, steady station running at full utilization with a queue in front of it is the constraint, even if nobody complains about it. The test is not noise or busyness. It is the queue and the starvation pattern: what is waiting, and what is waiting for it? Answer that, and the constraint names itself.
Free it, then find the next one
Freeing a constraint follows an order: first, make sure it never starves — feed it, staff it, keep its tooling ready, because an hour lost at the constraint is an hour lost for the whole line. Second, take work off it that does not belong there — a step another station could do, an inspection that could move. Third, only then consider capacity: overtime, equipment, or people. And when the constraint moves — as it will — find the next one and repeat. Constraints migrate, which is why bottleneck work is a routine, not a project. Reliability work belongs in the same conversation; see Total Productive Maintenance for keeping the constraint running, and The Lean Culture Blueprint for the system around it. Keep the constraint check in your pocket: https://apps.apple.com/us/app/lean-culture-advisory/id6790203662