The warehouse-factory boundary: where stock records stop matching reality

Two functions, one boundary, and no agreed moment of transfer

Ask a store keeper when material becomes production's responsibility and ask a supervisor the same question, and the answers rarely match. One says at picking, the other says at the machine. In the gap sits material that has left one set of records and not entered another, which is why it can be simultaneously issued and missing. Fixing this needs no technology: define a physical point, mark it, agree that crossing it is the transaction, and make one named role responsible for recording the crossing. Ambiguity here generates more variance than any amount of cycle counting will resolve.

Inbound staging: received, but not yet production's problem

Material arriving for a scheduled build often bypasses proper storage and goes straight to a staging area near the line, which is sensible and creates a familiar hazard. Staged stock is easy to raid for another job, easy to double-count, and easy to lose when the schedule changes and nobody returns it. Give staging a defined footprint, a maximum content expressed as jobs rather than volume, and a rule that anything removed for a different order is transacted. Without that, staging becomes a second uncontrolled store sitting in the most valuable space in the building.

Finished goods: the gap between made and put away

Output completed at the end of a line is not yet in stock, and the period before it is put away is where a surprising amount of damage, mixing and misdeclaration happens. Pallets built up over a shift, part-completed loads, items awaiting inspection release and rework sitting beside good product create a zone where anything can be picked up by anyone. Design the transfer point deliberately: a defined area, a clear physical distinction between released and unreleased output, and a booking event tied to the physical move rather than to the end of the shift.

Designing the interface as a place, not just a process

The physical arrangement determines how well the procedure survives contact with a busy day. Questions worth settling: whether inbound and outbound cross the boundary at the same point, how wide the transfer area is when both sides are at their peak, whether store trucks enter the production hall or stop at a line, how the area is lit and marked, and whether it is under the same supervision as either side. Where trucks and pedestrians converge at a doorway, traffic separation must be designed by people competent in workplace transport risk rather than improvised with tape.

Ownership, measures and the argument they create

Because both functions are measured on things the other side affects, the interface generates recurring friction: production says the store issues short, the store says production takes without booking. A workable arrangement treats the boundary as an internal service with agreed expectations on both sides, reviewed with the actual discrepancy data rather than with anecdotes. Publish the errors by type, not by department, so the conversation is about causes. Where the same discrepancy recurs, the cause is almost always a physical arrangement that makes the correct behaviour slower than the incorrect one.

Frequently asked questions

Should the store and production report to the same manager?

Common reporting removes the incentive to blame across the boundary and speeds up decisions, which is why many plants do it. The counter-argument is that an independent store protects record integrity, since the function recording stock is not the function under pressure to keep a line running. Either structure can work provided the transfer point is defined and the discrepancy data is visible to both sides. What does not work is separate reporting with no agreed interface, which converts every error into a dispute.

Why does our stock accuracy fall apart around the production area?

Usually because material physically moves faster than transactions do, and the procedure for recording a move is slower than the move itself. Typical causes are issues booked at end of shift rather than at the time, staged material taken for other jobs without record, returns to store that never get booked back, and scrap consumed but not declared. Look for the places where doing it correctly takes longer than doing it informally, and redesign those rather than retraining people who already know the rule.

How much space should the transfer area between store and plant have?

Enough to hold the peak, not the average, which is where most designs go wrong. Consider what accumulates when a delivery arrives at the same time as a shift's output is completed, and whether a stoppage on either side causes material to back up into the other. Undersized transfer areas spill into aisles, which is how the boundary becomes both a congestion point and a safety issue. Size it deliberately and defend it, because it is the first area anyone tries to reclaim.