How to stop losing tools on site without treating people like suspects

Ask any site manager where the missing SDS drill is and you will get a shrug, not an accusation. That shrug is the actual problem. Tools rarely vanish — they migrate. A van, a lock-up, a subcontractor who borrowed it for an hour in March, a toolbox belonging to someone who left in June.

This matters more than the replacement cost. It is the second drill bought because nobody could find the first, the hour lost looking, and the job that waited.

Why sign-out books fail

Every site has tried a book. They fail for a reason worth understanding before choosing a replacement, because most replacements fail the same way.

• Signing out is slower than not signing out, and nothing enforces it at the moment it matters.
• A book records that something went. It does not tell you what is still out, which is the question you actually have.
• Handovers are invisible. A signs it out, gives it to B, and the book still says A — so when you chase, you chase the wrong person and everyone learns the book is wrong.
• Nothing ever chases. The entry sits there for eight months and nobody reads a book backwards.

What works

Make the item identify itself

Every returnable tool gets a tag with a code on it. Not the serial number — a code you own, printed large, on a label that survives a site. Then taking it out is a scan rather than a description, which removes both the typing and the ambiguity about which of the four identical drills is out.

Make the person identify themselves in one action

Most people already carry an ID card. Scan its barcode or tap it if it has RFID. That is the whole identification step — no username, no password, nothing to forget, and it works with gloves on if you use RFID.

If cards are shared or non-existent, a typed staff number is a perfectly good fallback. What matters is that it takes one action, not that it is sophisticated.

Put a due date on it from the start

Not to be strict — to give the system something to chase. A tool that was due back last Tuesday can be emailed about automatically. A tool with no due date is not overdue, it is just gone, and nobody is ever prompted to ask about it.

A default loan period per category is usually enough: a day for consumable-ish hand tools, a week for plant, whatever matches how the work actually runs.

Show people what they are holding

The single highest-return feature in equipment tracking is showing someone, the moment they identify themselves for anything else, that they still have three items out and one is overdue.

Almost nobody is hoarding kit. They have forgotten, and no one has told them. A reminder at the moment they are standing in the store is worth more than a monthly report to a manager.

Ask about condition on the way back

The return is the only moment anybody will ever tell you a tool is broken. Ask then, in one tap, and route damaged kit to maintenance rather than back onto the shelf. Otherwise the next person discovers it, at the worst moment, and the loop never closes.

What makes it worse

1. Making it feel like surveillance. If the framing is "we do not trust you", people work around it and you lose the data entirely. The honest framing is true: this is so you are not blamed for something you returned, and so we stop buying things we already own.
2. Charging people for losses. It converts a recording system into a confession, and recording stops.
3. Tracking everything. Tag what is worth chasing. Putting a hammer through a custody workflow trains everyone to bypass the workflow.
4. A system only the storekeeper can operate. If kit can leave when they are on holiday, the record has a hole in it every August.

A realistic first week

1. List what is genuinely worth tracking. Usually far fewer items than expected — the expensive, the shared, and the statutory.
2. Tag them. An afternoon with a label printer.
3. Put a screen in the store room with a reader on it.
4. Use existing ID cards to identify people.
5. Do one amnesty: everyone brings back what they have, no questions, and it gets checked back in so the opening position is true.

That last step matters more than it sounds. Starting from a record you know is accurate is what makes the first overdue report credible — and the first credible overdue report is what makes people believe the system is real.