Changeover time is the only machine parameter that costs money without producing anything. Every minute of it is a minute of line capacity that can never be recovered, and on a multi-size line it is often the largest single source of lost output.

Most plants attack it by asking operators to work faster. That is the least productive available response. The productive response is to separate the work that requires the machine to be stopped from the work that does not — which is the entire content of the SMED method, developed for press changeovers and directly applicable here.

The Short Answer

A multi-size case erecting line with recipe memory and tool-free adjustments should change format in three to eight minutes. The same machine run without recipe memory typically takes twenty to forty-five minutes.

The gap between those two figures is not operator skill. It is the difference between adjustments that are set to a number and adjustments that are hunted to a position. Converting the second category into the first is a one-time engineering exercise that pays back on the next shift and every shift after it.

Four measures deliver most of the available saving, in this order: recipe memory for every motorised axis, tool-free and detented manual adjustments for everything not motorised, external preparation of the next format’s materials, and standardisation of the setting sheet so that the sequence is identical every time.

Where Changeover Time Actually Goes

Stopwatch a changeover before changing anything. The distribution is almost always the same and almost always surprising.

Finding the setting sheet typically consumes two to five minutes. The sheet is in a folder, the folder is in an office, and the previous operator’s handwriting is ambiguous.

Positioning the motorised adjustments consumes five to fifteen minutes if they are driven manually to an indicator, and under one minute if they are recalled from a stored recipe.

Adjusting the non-motorised points — magazine guides, suction cup positions, fold rail positions, glue nozzle position, discharge width — consumes ten to thirty minutes. This is the largest single block, and it is the block most resistant to improvement, because each point is independent.

Trial running and verification consumes three to ten minutes: two or three cases erected, measured, and the settings corrected.

Finding tools and parts — a spanner, an allen key, a shim — is scattered across the whole changeover and rarely measured, but it is where the repeated small delays hide.

The insight that follows from this distribution is that the largest block is the one nobody has engineered. Recipe memory handles the motorised axes, but most machines have more manual adjustment points than motorised ones, and those are usually left to operator judgement.

The Four Measures That Deliver

1. Recipe Memory on Every Motorised Axis

This is the highest-return single measure and the one to insist on at specification stage. Every servo axis on the machine — magazine height, fold rail width, compression height, discharge width — should be capable of recalling a stored position from the HMI by format number.

The practical questions to ask before buying are not “does it have recipes” but how many, can a recipe be edited in place, can a recipe be copied and modified, and can recipes be exported. A machine with thirty recipe slots and no export facility will run out of slots within two years on a growing product range. The recipe architecture differs between models, and the full product range shows which configurations carry full-axis memory.

2. Tool-Free Adjustment with Detents Everywhere Else

Every adjustment point that remains manual should be operable without a tool and should reference a scale or a detent position rather than an operator’s judgement.

A hand-wheel with a graduated scale and a lock is faster than a spanner on a bolt, because the operator reads a number instead of hunting a position. A guide with detented positions at fixed intervals is faster still, because the correct setting is a click rather than a measurement.

The specification to demand is: no tool required to change format, anywhere. A machine that needs a spanner during changeover will lose five to fifteen minutes per changeover to walking to the toolbox.

3. External Preparation

SMED’s central distinction is between internal work — which can only be done while the machine is stopped — and external work, which can be done while the machine is running.

On a case erecting line, external work includes fetching and staging the new board, checking the new blank against the setting sheet, pre-setting any quick-change carrier that can be prepared off-machine, and confirming that the recipe for the new format exists and is correct.

The gain is large and it is nearly free. If external work is moved out of the changeover window, the measured changeover time drops by the amount that was previously spent walking and looking. Two to six minutes per changeover is typical, and it requires only a staging trolley and a rule that the next job is prepared before the current job ends.

Where a machine uses quick-change carriers — a magazine insert, a suction cup plate, a fold rail cartridge — the carrier for the next format can be pre-set while the current format is running, and the changeover becomes a swap rather than an adjustment. This is the single largest available reduction on machines with high format counts, and it is the reason to specify interchangeable carriers at purchase rather than afterwards.

4. A Standard Setting Sheet and a Fixed Sequence

Changeover quality collapses when the sequence varies. Two operators changing the same machine to the same format in a different order will produce different times and, occasionally, different results.

A single-page setting sheet per format, listing every adjustment point in a fixed order with the target value, removes the variation. The sheet should carry the values the machine was validated at, not the values from the last casual adjustment.

The sheet also becomes the training document. A new operator with a good setting sheet and a detented machine reaches competent changeover in days rather than weeks.

Measuring the Improvement

The measurement is trivial and it is what makes the improvement real. Record three numbers per changeover: start time, first good case out, and the format number. From those, compute:

Changeover time = first good case out − machine stopped
Changeover cost = changeover time (hours) × line contribution margin per hour
Annual changeover cost = changeover cost × changeovers per year

On a line running four changeovers per shift across three shifts, that is roughly 3,000 changeovers per year. Reducing changeover from thirty minutes to six minutes at a line contribution of USD 120 per hour saves about 1,440 hours of line capacity annually — which is not a soft saving but a recoverable capacity that either absorbs growth or avoids a second machine. Where the format range is wide but volumes are moderate, a Single-Piece Carton Erector configured with interchangeable carriers usually reaches a lower changeover time than a general-purpose machine, because each format gets tooling rather than an adjustment.

What Not to Optimise

Three things look like changeover savings and are not.

Skipping the trial run. Erecting two cases and measuring them costs ninety seconds. Skipping it risks a jam downstream that costs thirty minutes. The trial run is cheap production, not waste.

Loosening the adjustment locks. A guide that is not locked will move during running. The result is an intermittent fault that takes longer to find than the changeover ever saved.

Setting the machine slightly oversize to save adjustment. Running a 400 mm case in a 410 mm channel reduces the adjustment time and increases the jam rate. Any saving is consumed by micro-stoppages within the shift.

Machine Settings Relevant to Frequent Changeover

The narrow bands in this table are the reason detents and scales matter. A 2 mm band cannot be hit by feel reliably, and it cannot be hit fast by feel at all.

Alarms That Reveal an Incomplete Changeover

E-29 Blank Feed Timeout appearing within the first fifty cases after a changeover is the signature of an incomplete magazine adjustment. It is worth logging changeover completion against first-alarm timing, because a format that reliably alarms within fifty cases has a setting sheet that is wrong.

Maintenance Rhythm for High-Frequency Changeover

Every shift — confirm the setting sheet for the next format is staged, check that all adjustment locks are engaged, and record the actual changeover time against target.

Every week — verify that the stored recipes still produce a square case on trial, check guide scales and detents for damage, and inspect quick-change carriers for wear.

Every month — review the three slowest changeovers of the month and identify whether the delay was a machine limitation or a preparation failure. The answer determines where the next improvement should go.

Every quarter — audit recipe accuracy across the full format range, re-verify the narrow-tolerance points against measured cases, and review whether any format needs a carrier rather than an adjustment.

Annually — re-baseline changeover time across the format range, replace worn detent mechanisms and scales, and re-issue setting sheets from validated values rather than from accumulated corrections. Setting sheet templates and validated value ranges for common formats are available through our support section.

FAQ

Is three minutes of changeover realistic?
On a machine with full recipe memory, tool-free adjustments and quick-change carriers, yes — three to five minutes is achievable. Without those features, no amount of practice gets below fifteen.

How many recipe slots do I need?
Count current formats, multiply by two for the range growth most plants experience within three years, and add ten spare. Export capability matters more than slot count, because it allows off-machine backup and editing.

Does changeover speed matter if I only change once a week?
Then it does not. Changeover engineering pays where changeovers are frequent — more than two per shift. Below that, spend the effort on reliability instead. High-volume single-format lines are better served by the Double-Head High-Speed Case Former, where the format barely changes and the priority is raw case rate.

What is the single best improvement?
Moving preparation out of the changeover window. It costs almost nothing and typically removes two to six minutes immediately, before any capital is spent.

Looking to cut changeover time on a multi-format line? Send your format list, blank dimensions and current changeover time to [email protected] or WhatsApp +86 13681839278. Our engineers will return a specification with recipe capacity, tool-free adjustment points and a carrier recommendation within 24 hours.