Warehouse location labels: specifying a label that is installed and stays installed

01 / FIELD NOTE

Keep the decision tied to the operating context.

A carton label and a shelf label look like the same object and are not. The carton label is applied at packing, read a handful of times as the carton moves, and destroyed with the packaging. The shelf label is applied once to a fixed position and is then read every time anyone works near it, for as long as that position exists. The second of those is a piece of installed equipment, and specifying it as though it were the first is the mistake this note is about.

The surface decides more than any other single factor. Shelving is painted steel, galvanised steel, wire mesh, plastic or laminate, and each behaves differently under a tag. Metal is the difficult case because it reflects the signal and detunes an inlay placed directly against it; the usual remedies are a spacer that holds the inlay away from the surface or an inlay designed for on-metal use, and both are choices made at order time rather than at installation.

Related to the surface is the question of what the label is attached to and how that object is used. A label on the front edge of a shelf is read from the aisle at a shallow angle. A label on the shelf deck is read from above, and may be covered by whatever is standing on it. A label on the upright is read from the side and is the one most likely to be struck by a forklift or a pallet. Those are three different read geometries, and the same inlay will not perform equally in all three.

Read distance and reader angle have to be established in place rather than on a bench. The relevant questions are how far the reader will be when it reads this label, from which direction, at what power, and whether the labels on the next bay along will also answer. A test on a desk establishes that the inlay works; it does not establish that this installation works, because none of the reflections, obstructions or neighbouring labels that decide the answer are present on a desk.

Cross-reads deserve particular attention on shelving, because the failure is asymmetric. Reading a label in the next bay is not a dramatic failure — it produces a plausible location that happens to be wrong, and a wrong location is worse than no location, because it will be acted on. Spacing the labels, choosing the read direction, and setting the reader power so that the intended label is inside the field and its neighbours are not is the work that prevents it.

The encoding has to be settled before the labels are printed, not after. A location label carries an identifier that the warehouse system already understands, or it carries one that something has to translate, and the translation is where the two drift apart. The questions worth answering early are whether the identifier is the location code itself or a serial that maps to it, whether the same scheme will be used at other sites, and what happens to the label when a bay is re-labelled or a location is split.

Durability is the property that separates a location label from every other kind. It has to keep its adhesive through the temperature range the building actually reaches, including any chilled area; keep its printed face readable through wiping, abrasion from passing goods and cleaning agents; and resist edge lifting, which is how most labels fail — not by falling off but by lifting at one corner until the inlay is damaged or the printed code becomes ambiguous.

Those requirements point at a synthetic face material and an adhesive matched to the surface and the temperature range, which is a different specification from the paper facestock that suits a carton. Where the printed code is used as a visual fallback by people who will not be carrying a reader, the print method has to survive the same conditions the inlay does, because a legible label with a dead inlay and an unreadable label with a working inlay are both single points of failure.

Expansion is worth deciding before the first order. A scheme that covers the current bays but has no room for a new aisle, a new zone prefix or a second site will need re-labelling, and re-labelling is expensive precisely because the labels are installed. Leaving space in the identifier scheme costs nothing at design time and is close to impossible to add later without touching every shelf.

The combination of a printed code and an inlay on one label is common and worth deciding deliberately. It lets a person read the location without equipment and lets the system read it without a person, and the two can disagree — which is useful during a transition and ambiguous afterwards. Deciding which of the two is authoritative, and what a disagreement means, is part of the specification rather than an implementation detail.

All of this is cheaper to establish by walking the building than by reading a datasheet. A short survey that records, for each type of bay, the surface material, the read direction, the expected reader and the neighbouring labels is what turns the specification into something a supplier can quote against — and what stops the surprise from arriving after the labels are on the shelves, when changing them means visiting every location again.