We run our tool crib on a spreadsheet. What do we actually gain by moving to a tool vending machine?

You gain a record that is created by the act of issuing rather than by somebody remembering to write it down. A spreadsheet crib is accurate only as long as every issue is typed in, and in practice the entries stop on the busy shifts, at night, and whenever the attendant is away from the counter — which is exactly when consumption is highest. A controlled dispense point authenticates the worker, shows only the items they are entitled to draw, dispenses the item, and writes the transaction with employee, item, quantity, job ID, machine, location and timestamp. From that record you get real stock counts instead of estimates, consumption attributed to a job or cost centre instead of a single tooling line, and replenishment triggered by an alert instead of by somebody noticing an empty shelf. The spreadsheet is not the problem; the manual data entry the spreadsheet depends on is.

What does a manual tool crib actually cost, and how do I build the baseline?

Four costs, and only the first one appears in the tooling budget. First, purchases: what you spend annually on tooling and consumables issued through the crib. Second, labour: attendant hours spent handing out items and the hours anyone spends counting, reconciling and chasing paperwork, multiplied by a loaded hourly rate. Third, expediting: the premium paid on emergency orders and urgent freight when an item runs out mid-job. Fourth, unrecorded issue — hoarding, personal stashes and tools that never come back for re-grind, which shows up as a gap between what you buy and what production can account for. Build the baseline from your own purchase ledger, your own attendant hours and your own rates. Then work out what one percentage point of reduction is worth, because that is the number the business case turns on. Any supplier who quotes a saving before seeing those figures has calculated nothing about your site.

Which items should go into the machine first?

The items with the highest annual value multiplied by the highest issue frequency — typically carbide inserts, drills and end mills in a machine shop, abrasives and welding consumables in fabrication, and gloves and general consumables in maintenance. High-turn items generate the transaction volume that makes the data useful within weeks rather than years, and they are where uncontrolled issue costs the most. Do not attempt to automate the whole crib at once. Automate the top items by spend and movement, run them long enough to see real consumption, tighten par levels against that data, and then decide whether a second tranche is worth a second machine. Items you issue a handful of times a year belong on a shelf, not in a coil.

Can the machine track tools against a job or work order number?

Yes — job or work-order capture is the point of tool vending as opposed to PPE vending. The operator selects or enters the job, work order or cost centre at the point of issue, and the job ID is written to the transaction record along with the employee, item, quantity, machine, location and time. That turns tooling from a single monthly overhead into a cost attributable to the job that consumed it. Pushing that record into an ERP, CMMS or tool management system is supported through the VoT cloud platform via REST API, CSV export and import, or a direct database connector, and it is scoped as a project against your system rather than shipped as a preconfigured module — send us the system you run and we will assess the integration specifically.

What about expensive gauges, micrometers and instruments that have to come back?

Those belong in locker cells rather than coils. A locker cell issues the instrument to an identified worker and accepts the same item back into the same cell, so the system knows what is out and who holds it, and the next issue can be made conditional on the previous one being returned. Optional per-cell weight sensors verify that the returned item is physically present rather than that the door was opened and closed. Calibration remains a human process — the machine records custody and return, but it does not inspect the instrument or judge whether it is still within calibration, so the calibration schedule stays with whoever owns it today.

When is a tool vending machine the wrong answer for a crib?

When the item moves too slowly, is too large, or is bought for one job. A cell occupied by an item issued twice a year earns nothing; that item is better left on a shelf with a periodic count. Long bar stock, large toolholders, awkward assemblies and anything that will not fit a coil, tray or cell is a dimensional problem no software fixes. Special tooling bought per project is not stock and should not be treated as stock. And where the crib attendant's real function is advising which insert grade suits the material, that is engineering judgement, not handing out boxes — automating the counter does not replace it, and a site that removes the advice along with the counter has made things worse. The honest scope is usually a machine for the fast-moving consumables and a much smaller, better-organised crib for everything else.

Tool Crib Modernisation — From Spreadsheet to Tool Vending

Modernising a tool crib is an inventory and entitlement exercise before it is a hardware purchase — baseline what the crib costs today, then automate the fast-moving consumables first and leave the rest on the shelf. Start by measuring the four costs a spreadsheet crib hides: attendant hours, unrecorded issues, shrinkage and hoarding, and expedited orders caused by stockouts. Inventory each item you intend to automate with its pack size, physical dimensions, consumption rate, unit cost and who is entitled to draw it, because that list decides the machine configuration. Migrate in phases — highest-turn consumables into coils first, job or work-order capture next, returnable instruments into locker cells after that, and system integration last. Expect the sign-out sheet to be replaced by a transaction record carrying employee, item, quantity, job ID, machine and timestamp. Low-turn, oversized and one-off tooling should stay in the crib; forcing it into a machine costs more than the control is worth.

Who this page is for

You are running a tool crib on a spreadsheet, a paper sign-out sheet, or an attendant’s memory. It works well enough on a quiet day shift and it stops working on a busy one. Stock counts are estimates. Job costing for tooling is a single monthly figure that nobody can break down. Somebody occasionally discovers a drawer of personal stock in a toolbox, and every few weeks a job stops because a common insert has run out.

This page is about moving that crib to a controlled dispense point without pretending the exercise is simple. It covers what the current arrangement costs, what to inventory before anyone quotes you a machine, how to phase the change, what the data lets you do afterwards, and — importantly — which parts of the crib should not be automated at all.

If you want the machine specification rather than the migration, that is on the tool and parts vending page.

What a manual crib costs

Only one of these four costs appears in the tooling budget.

Cost	Where it hides	How to measure it
Purchases	The tooling and consumables line	Annual spend on items issued through the crib, from the purchase ledger
Crib labour	Wages, not tooling	Attendant hours at the counter, plus hours anyone spends counting, reconciling and chasing sign-out sheets, at a loaded hourly rate
Expediting	Freight and premium pricing	The extra paid on emergency orders and urgent freight when an item runs out mid-job
Unrecorded issue	Nowhere — it is a gap, not a line	The difference between what you buy and what production can account for: hoarding, personal stashes, tools never returned for re-grind

An unstaffed crib does not remove the labour cost, it redistributes it. Every operator who walks to an open crib, searches for an item and walks back is spending production time on stores work, and no record is produced at the end of it.

Baseline before you buy

annual crib cost = tooling and consumables purchased + (attendant and reconciliation hours × loaded hourly rate) + expediting premium

Run that with your own ledger, your own hours and your own rates. Then calculate what one percentage point off the total is worth. That single figure decides whether the project is worth doing, and it is the only honest starting point for a payback discussion. Published reference ranges exist — sites moving from open cribs to controlled dispensing commonly report tool and consumable reductions in the 25–40% band — but a range measured elsewhere is context for your estimate, not a forecast for your site.

The same reasoning applied to safety equipment is set out on the PPE cost control page, which includes a fully worked illustrative baseline you can copy the arithmetic from.

What to inventory before you automate

The inventory list is the specification. A supplier cannot configure coils, trays and cells without it, and a list assembled properly is useful even if you decide not to buy anything.

For every item you are considering automating, record:

Field	Why it matters
Part number and description	The identity the system, the ERP and the purchase order all have to agree on
Pack or issue unit	Whether you issue a single insert, a box of ten, or a sleeve — the dispense unit must match how people actually draw it
Physical dimensions and weight	Decides coil pitch, tray or cell; this is what makes an item dispensable or not
Annual consumption and issue frequency	Ranks the migration and sets par levels
Unit cost	Ranks the migration by value at risk
Who may draw it	The entitlement rule: by role, department, shift or ticket
Consumable or returnable	Consumables go to coils and trays; returnable instruments go to locker cells
Job costing required	Whether issues must carry a job, work order or cost centre reference

Two practical notes. First, the dimensional data is the part teams skip and the part that causes dispense failures later — check product-to-mechanism fit before build using the planogram configurator and the coil options, not after delivery. Second, this exercise almost always finds duplicate part numbers, obsolete stock and items nobody has issued in two years. Clean that up first; automating a messy catalogue preserves the mess and pays to store it.

Phasing the migration

Do it in stages. Each stage produces something usable on its own, and each one informs the next.

Phase	What happens	What you get at the end
0 — Baseline and inventory	Measure the four costs; build the item list above; clean the catalogue	A defensible business case and a machine specification
1 — Highest-value consumables	Automate the top items by annual spend × issue frequency; identity-based issue turned on	Real stock counts and attributed consumption on the items that cost the most
2 — Job and entitlement capture	Job, work order or cost centre selected at issue; entitlements set by role and department	Tooling costed to jobs; issue restricted to who should be drawing it
3 — Returnable and high-value items	Instruments and gauges moved to locker cells with return enforcement	Custody visible: what is out, who holds it, what came back
4 — Integration and tuning	Push transactions to the ERP, CMMS or tool management system; reset par levels from real data	Replenishment driven by consumption rather than by noticing an empty shelf

Run phase 1 with generous limits. The first period of real data is worth more than a carefully guessed configuration, and limits set too tight on day one turn the machine into an obstacle that operators route around — which produces exactly the hoarding the project was meant to remove.

Keep the spreadsheet running in parallel through phase 1 for the automated items only. Where the two disagree, the disagreement is the finding.

What sign-out looks like afterwards

The operator identifies themselves with an RFID employee card or a PIN on the machine’s reader.
The touchscreen shows only the items that operator is entitled to draw, by department, role and shift.
They select the job, work order or cost centre the issue belongs to.
The item is dispensed from a coil, tray or locker cell.
The transaction is written with employee name, key card number, time, product, quantity, machine name, location and job ID.
Stock level updates in real time on the cloud dashboard, and a low-stock alert fires when the item hits its reorder point.

Nobody has to remember to record anything, which is the entire point.

What data comes back

The value of the migration accrues in the transaction history, not in the steel.

What the record shows	What you can decide with it
Consumption per item per period	Par levels and reorder points based on measured demand
Consumption per job or work order	True tooling cost per job, and which jobs are outliers
Consumption per operator or cell	Where tool life differs on identical work — a process question, not an accusation
Issue frequency by hour and shift	Whether the crib’s opening hours were the real constraint
Stock on hand per machine, live	Counting time removed; stockouts predicted rather than discovered
Custody of returnable instruments	What is out, who has it, and what has not come back
Refusals against entitlement limits	Where a limit is wrong, or where an item is wrong for the people drawing it

That last row matters. Repeated refusals on one item are a signal to investigate, not a saving to record — the same principle that governs quota design on the PPE cost control page.

Optional per-cell weight sensors add gram-level verification of what left the cell, which is useful for loose items such as fasteners where a coil count and a physical count drift apart.

Where vending is the wrong answer

A page that recommended automating the whole crib would be selling, not advising. Leave these where they are:

Low-turn tooling. An item issued a few times a year does not justify a cell. Shelf it and count it periodically.
Oversized and awkward items. Long bar stock, large toolholders, ladders and bulky assemblies are a dimensional problem, and no amount of software solves a fit problem.
One-off and project-specific tooling. Special tooling bought for a single job is not stock and should not be managed as stock.
Items whose real constraint is advice. Where the attendant’s function is knowing which insert grade suits the material, that is engineering judgement. Automating the counter does not replace it, and removing the advice along with the counter makes the shop worse.
Calibration workflows. A locker cell records custody and return of a micrometer. It does not inspect it or judge whether it remains in calibration; that schedule stays with whoever owns it now.
Sites too small to generate data. If the whole crib is a cupboard and issue volume is low, the transaction history will never be dense enough to act on. Better shelving and a discipline about the sign-out sheet may be the correct answer, and it costs less.

The realistic end state at most sites is a machine holding the fast-moving, high-value consumables and a smaller, tidier crib holding everything else.

Questions to send any supplier

Copy these into your enquiry. They separate a system that controls issue from a box that drops things.

How does an operator identify themselves, and will it work with the employee cards we already issue?
Can a job, work order or cost centre be captured at the point of issue, and is it stored on the transaction?
Can entitlements be set per item, per role and per period, or is there only one global limit?
Can a supervisor override a limit, and is the override recorded?
Which of our items fit a coil, which need a tray, and which need a locker cell — assessed against our dimensions before build?
Can the same cell issue an item and accept its return, and can the next issue be blocked until it comes back?
What fields are written on each transaction, and how do we export them?
What happens to dispensing and logging if the site network drops?
What triggers a replenishment alert, and can par levels be changed from the dashboard?
Who owns the transaction data, and how do we get it out if we change supplier?

Question 10 belongs with everything in the manufacturer vetting guide. The history this system builds is the asset; it matters who controls it.

What KioskForce does on each point

Point	Our answer
Operator identification	RFID employee card or PIN on the machine’s integrated reader
Job costing	Job, work order or cost centre selected at issue and written to the transaction
Entitlement	Touchscreen shows only the items that operator may draw, by department, role and shift
Dispense mechanisms	Coils, trays and locker cells mixed within one machine, configured to your item list
Product fit	Checked against your dimensions before build, using the planogram configurator and coil options
Returnable instruments	Hybrid locker cells issue and accept returns in the same cell; next issue can require the return
Verification	Optional per-cell weight sensors for gram-level confirmation of what left the cell
Transaction record	Employee name, key card number, time, product, machine, location, job ID
Stock visibility	Real-time levels per machine with low-stock alerts on the cloud dashboard
Integration	ERP, CMMS and tool management systems via REST API, CSV export and import, or database connector — scoped per project against your system
Savings forecast	Not quoted before we see your consumption data. The baseline above is yours to run

Hardware and software are both designed in-house, so the dispense mechanism, the entitlement rules and the reporting are specified together rather than assembled from three vendors. See customisation for how a configuration is agreed, and spare parts and support for what happens after it is installed.

Tool and parts vending machines — the machine: coils, trays, locker add-ons, and what it dispenses.
Industrial vending hub — PPE, tool and parts, and point-of-use MRO on one platform.
Point-of-use MRO vending — consumables issued at the production cell rather than at a central store.
PPE cost control and quota-based issue — the same control model applied to safety equipment, with a worked baseline.
Per-cell weight sensor tracking — verifying what actually left the cell.
Planogram configurator and coil options — checking product fit before the machine is built.
Work-order kitting and BOM dispensing — resolving a work order reference into the kit of parts that job needs.
ServiceNow integration — the integration pattern, applied to IT asset dispensing.
Manufacturing industry solutions — the same problem in its usual site context.
How to vet a vending machine manufacturer — before you commit to a supplier.
Customer Service FAQ — MOQ, lead times, warranty and support in short form.

Get started

Send the item list described above — part numbers, pack sizes, dimensions, annual consumption and who may draw each item — with your headcount and the systems you need the data to reach. We will tell you which items are dispensable, which are not, and what a first-phase configuration looks like.

Contact Us for More Information