The Broken Sensor That Cost a Day
A proximity switch failed on a packaging line at 2 a.m. on a Saturday. The maintenance tech called the supplier — nobody answered. The line was down until Monday. A $20 sensor caused $40,000 in lost production. The root cause was not the supplier. It was the spares room: there was no spare of that sensor on the shelf. Worse, when we finally looked, the spares room held 47 relays of a type the plant had not used in five years. We were stocking the wrong things.
Spare parts inventory is a trade-off: too little and you wait for deliveries; too much and you tie up cash in obsolete parts. Here is how to find the middle.
The ABC Method: Classify by Downtime Cost
Not all parts are equal. Sort them by the cost of having them fail:
- A-class (critical, high downtime cost): A servo drive, a PLC CPU, a gearbox, a main motor. If these fail, the whole line stops. Stock at least one spare on the shelf, even if the part costs $2000.
- B-class (important, replaceable): A proximity sensor, a pneumatic valve, a relay, a fuse. These fail often and cause localized downtime. Stock 2–5 spares.
- C-class (cheap, long-life): Fasteners, labels, cable ties, fittings. Stock in bulk (a bin). Nobody needs a single spare screw — you have a box.
The mistake most plants make is the opposite: they stock C-class (screws, relays) in abundance, and skip A-class (a $2000 servo drive) because “we can get it in two days.” Two days is exactly what you cannot afford when the line is down.
Lead Time: The Hidden Number
Before you decide how many spares to stock, find the supplier’s lead time. A sensor from a local distributor is 24 hours. A custom servo cable from overseas is 8 weeks. The rule: stock enough to cover the lead time plus a safety margin. If the lead time is 8 weeks, you need a spare on the shelf — no argument. If it is 24 hours, a spare is nice but not mandatory.
The inventory rule: Stock A-class by lead time (cover 2× lead time). Stock B-class by failure rate (2–5 spares). Stock C-class in bulk. Throw out parts that have not moved in two years (they are obsolete).
The Dead Stock Problem
Spare rooms accumulate parts that nobody needs. A machine is replaced; the old parts sit on the shelf. Every year, take a walk through the spares room. Mark anything with no label, no known machine, or no demand in 24 months. Sell it, scrap it, or return it. A cluttered spares room is a tax on the maintenance budget.
When to Buy a Spare — Even Before the Failure
There is a counter-intuitive move that pays off: when you install a new machine, buy one of every A-class part at the same time (the manufacturer often gives a “spare package” at discount). The machine is new, but the day will come when the drive fails, and the 20% discount at purchase beats the emergency freight surcharge at 2 a.m.
Spare Levels by Example
| Part | Class | Lead Time | Spares on Shelf |
|---|---|---|---|
| Servo drive (main axis) | A | 4 weeks | 1 |
| PLC CPU | A | 2 weeks | 1 |
| Proximity sensor (common type) | B | 24 hours | 5 |
| Pneumatic valve | B | 48 hours | 3 |
| Fuse 5A | C | Same day | 20 (bulk bin) |
| Timing belt (specific) | A | 6 weeks | 2 |
Track the Spares That Actually Failed
Keep a simple log: every part that failed in the last 12 months. Sort by frequency. The parts that failed three times go up a class. The parts that never failed go down. This is how the spares room evolves from guesswork to data.
A Spares Audit Checklist
- List every A-class part on every critical machine.
- For each, check lead time from the supplier.
- Stock 1× lead time (minimum 1) on the shelf.
- Review failures from the last 12 months; bump frequency parts up.
- Scrap anything with no demand in 24 months.
- Label every bin with part number, machine, and quantity.
- Buy the A-class spares with the machine purchase (discount).
- Reorder when stock hits the reorder point (not zero).
The Bottom Line
The $20 sensor that cost a day exposed a spares room stocked for the wrong threats. Classify by downtime cost, cover the long-lead parts, and clear out the dead stock. The goal is not a full shelf — it is the right shelf. When the 2 a.m. failure hits, the part you need is the one you decided to keep.
