Changeover time is the tax every multi-format line pays. The machines matter, but the habits around them decide whether a format change costs ten minutes or ninety.
Make settings numeric, then trust them
Every handwheel with a position counter, every servo axis with a stored recipe, is a setting that restores without trial runs. The discipline is writing the numbers down once, verifying them once, and then refusing to “just tweak it” without updating the recipe.
Stage format parts like a pit stop
A changeover that starts with searching for parts has already failed. A shadow-board trolley per format, staged at the machine before the last good pack of the outgoing run, cuts the clock more than any mechanical upgrade at similar cost.
Split internal from external work
Anything that can be done while the machine still runs — heating the next glue program, loading the next film reel on the second mandrel, pre-counting cartons — should never happen during the stop. Most lines find a third of their changeover hiding in this one distinction.
Time it, every time
Record every changeover: format pair, duration, first-pass yield after restart. The pattern tells you which format pairs deserve engineering attention and which need practice. What gets measured gets shorter.
