Quick changeover method: dividing the work that needs the machine stopped

Record the changeover before anybody argues about it

Film the whole thing, from the last good piece of the previous run to the first good piece of the next, with the camera positioned to see the operator's movement rather than just the machine. Then watch it with the setters, the supervisor and an engineer, listing every element and how long it took. The recording matters because memory is unreliable and because it removes the argument about whether something normally happens. What surprises most teams is how much of the elapsed time is spent searching, walking, waiting for a crane or fetching something that should have been at the machine.

Internal and external is the distinction that does the work

Sort every listed element into work that physically requires the equipment stopped and work that could be done while it runs. Applied honestly to a poor changeover, a large share of the internal list turns out to be external work being done at the wrong time: collecting tooling from a store, locating the setting sheet, finding the right fasteners, waiting for material to arrive, cleaning containers, hunting for a trolley. None of that requires a stationary machine. Simply moving those elements outside the stopped period, before any equipment is modified, generally produces the first and least costly reduction.

Converting internal work into external work

The next stage costs money and produces the larger gains. Duplicate tooling so the next set can be pre-assembled offline. Pre-heat dies and pre-set tools on an offline fixture with a gauge. Kit the changeover so every fastener, insert and consumable arrives on one trolley with a checklist. Stage the first material and the paperwork at the machine before the previous run ends. Each of these converts a task that used to stop production into preparation done in parallel. This is where the method typically stalls, because the fixture or the second tool set needs approval that nobody has requested.

Streamlining what genuinely has to stay internal

For the remainder, attack fastening and adjustment. Replace threaded fasteners with quick-release clamps or standardise on a single fastener size so one tool serves. Use locating pins, stops and shims so position is repeatable rather than measured. Eliminate trial-and-adjust cycles by making the correct setting mechanical instead of judged, since adjustment and first-off correction often consume more time than the physical change. Where two people can work in parallel, define who does what and in what order, and be explicit about the isolation and safe-access rules that apply when several people work on one machine.

Why the improvement decays and how the plant knows

Three failures recur. The fixture or duplicate tooling identified during the exercise is never funded, so the analysis is filed and the changeover reverts. The new sequence is demonstrated by the best setter and never written down, so it lasts only while that individual is on shift. Or the changeover time falls and nothing downstream changes, because the plant continues to run the same long batches, meaning the benefit was captured as extra idle capacity rather than as shorter lead time. Track the actual duration by shift and by setter, and watch whether run lengths have moved at all.

Frequently asked questions

Which machine should we apply the method to first?

Start where changeover time actually costs the plant something, which usually means the resource that limits output, or a machine whose long setup forces batches larger than the schedule wants. Applying the technique to equipment with spare capacity produces a satisfying reduction and no benefit, because the hours released were never needed. Confirm the choice by asking what the plant would do with the recovered time, and if there is no answer, choose a different machine.

Do we need to buy duplicate tooling to make this work?

Not for the first stage. Reorganising so that preparation happens while the machine runs, kitting the tools and materials, and putting the setting information at the machine all cost very little and are often the largest single step from a genuinely poor starting position. Duplicate tooling, offline pre-setting fixtures and quick-clamping become the next stage, and they should be justified by what the recovered capacity or the shorter runs are worth, not by the reduction on its own.

How is this different from managing changeovers day to day?

This is the analytical exercise that produces a better method: observe, classify, convert, streamline, write the new standard. Running changeovers well from shift to shift is a separate operational discipline concerned with scheduling them sensibly, sequencing to minimise the amount of change, having the setter and material ready, and recovering when one overruns. The method creates the capability; daily management determines whether the plant actually gets the benefit.