Kitting for production: when a pre-picked part set is worth the extra handling

Kitting relocates the picking problem, it does not delete it

Somebody still walks to every bin and counts every item; the question is whether that happens at the workstation during production time or beforehand in a preparation area. Moving it off the line frees station space, cuts operator walking and lets one prepared set serve a whole build. It also adds a handling step, a container, a staging location and a new opportunity for error. The honest comparison is total labour and error rate across both arrangements, not the visible tidiness of the line, which improves either way and tells you nothing about cost.

The conditions under which a kit earns its keep

Kits pay where variety is high and stations cannot physically present every component, where parts are small and numerous enough that picking dominates the cycle, where a build must be traceable to a specific set of components, and where a job moves between several stations that would each otherwise need the same items. They rarely pay on stable, low-variety, high-volume work with a handful of parts, where direct line-side supply is simpler and faster. Mixed answers are common in one building: kit the variable content, feed the common fasteners and consumables straight to the station.

Kit accuracy is the entire proposition

A kit the builder does not trust is worse than no kit, because the operator opens it, finds a shortage halfway through the build and now has a part-finished job blocking a station. Accuracy needs a verifiable pick list tied to the correct drawing revision, a count check on items where miscounts matter, and a clear marking convention for a kit that is deliberately incomplete. Track short kits as their own measure with the cause recorded — stock error, substitution, revision mismatch, picker error — because the four have entirely different fixes and get lumped together by default.

Where, when and against what the kit is built

Kits built too early tie up parts, occupy staging space and go stale when the order changes or an engineering change lands. Built too late, they become the reason the line is waiting. Preparation should run a defined interval ahead of the schedule, against the released works order and its current revision, in a marked location with kits addressed to the job rather than stacked generically. Physical staging discipline matters more than it sounds: unlabelled kits on a common rack get cannibalised for a hot order, and nobody records which one lost its parts.

Returns, breakage and closing the loop

Kits come back. Surplus items from cancelled orders, parts damaged during assembly, spares deliberately included for delicate components — all of it has to return to stock with the record adjusted, or inventory accuracy decays quietly and every subsequent kit gets harder to pick. Set a rule for what happens to a returned kit within the shift, name the person who books it back, and check the exceptions. Where breakage during build is routine, add the replacement quantity to the kit deliberately and record the consumption; letting operators quietly pull extras from a nearby bin destroys both the count and the traceability the kit was meant to provide.

Frequently asked questions

Who should check a kit before it goes to the line?

The picker should verify their own count against the list, and a second check is justified only where a shortage would be expensive to discover mid-build — long jobs, controlled assemblies, or components with long replacement lead times. Blanket double-checking every kit usually costs more than the errors it catches. A better investment is making the pick itself hard to get wrong: clear locations, unambiguous part identification, and a list that matches the drawing revision in force.

Can we kit and still run engineering changes cleanly?

Yes, provided the effectivity point is defined and kits carry the revision they were built to. The danger is kits prepared before a change and consumed after it, so the build silently uses superseded parts. Set a rule that a kit is picked against the revision current at release, and that any change with immediate effectivity triggers a review of kits already staged. Mark and quarantine the affected ones rather than trusting anyone to remember.

How far ahead should kits be prepared?

Far enough that preparation is not a bottleneck on the day, close enough that the order is unlikely to change. Most plants land on roughly the length of their frozen schedule window, so kits are built for work already committed. Staging space usually settles the argument — if the kit rack is overflowing, preparation is running too far ahead and is absorbing parts that other orders may need more urgently.