Building a Proximity Sensor Spares List for Your Plant

Most plants we deal with hold somewhere between thirty and sixty proximity sensor part numbers, one or two of each, accumulated over years from whoever supplied the machine. The result is predictable: a large amount of capital on the shelf, and the specific sensor that failed at two in the morning is not among it.

Rationalising that holding is usually worth more than any individual specification decision, and it is an exercise you can run in an afternoon.

Why the holding grows the way it does

Nobody designs it. Each machine arrives with its builder’s preferred sensors, each breakdown gets solved with whatever was available, and each engineer has a brand they trust. Nothing is wrong at any single step, and after ten years you have forty part numbers covering roughly eight genuine requirements.

Step one: list what is actually installed

Walk the plant, or pull the maintenance records, and record for each sensor position: the part number, the body size, the sensing distance, the output type and function, the target material, and the machine it is on.

Two things usually emerge immediately. A large share of positions are doing an identical job with different part numbers. And several sensors are substantially over-specified for what they do, usually because they were specified by a machine builder covering themselves.

Step two: group by requirement, not by part number

Sort the list by what each sensor has to do rather than by what is fitted. Target material, sensing distance needed, body size the mounting allows, output the input card expects, environment.

Most plants find that eight to twelve genuine requirements cover everything. That is the number of part numbers you should be holding.

Step three: pick one part per group, specified for the worst case

For each group, choose a single part that covers every position in it. Specify it against the most demanding position in the group, not the average. A sensor rated for washdown will work perfectly well in a dry location, but not the reverse.

Two rules make this work:

Go up in body size where the mounting allows. An M30 with 15 mm sensing covers more positions than an M18 with 5 mm, because the extra standoff absorbs dust, scale and mechanical slop. It costs slightly more and eliminates several part numbers.

Standardise the output. If most of the plant is PNP, make the exceptions the thing you change rather than carrying both. Where a machine genuinely needs NPN, that is one part number, not a parallel range.

Step four: decide the quantities

Stock against consequence, not against count. Three questions per group:

• How many positions does it cover? More positions means more chance of needing one.
• What happens when one fails? A sensor on a safety interlock or a production bottleneck is worth holding deeper than one on a non-critical conveyor.
• How quickly can it be replaced? A part available locally in a day needs less on the shelf than one on a six week import lead time. This alone is an argument for domestic sourcing on anything critical.

Step five: the things people forget to stock

The sensor is rarely the only thing that fails.

• Locknuts in the right thread pitch. M18 proximity sensors use M18 x 1 fine, and an ordinary hardware shop will hand you coarse.
• Mounting brackets, particularly for positions where the bracket gets hit.
• Cordsets if any sensors use connectors, because the cordset fails as often as the sensor.
• Control units and amplifiers for NAMUR loops. A spare sensor is useless if the amplifier failed.
• Reflectors for retro-reflective photoelectric positions.

Step six: write it down where it will be found

A rationalised list that lives in one person’s head stops working when they are on leave. Record for each stocked part: the part number, what it replaces, which positions it covers, the minimum stock level, and the supplier with a contact. Put the Accent part number on the machine drawing as well as on the stores label, so the next engineer does not have to work it out.

What it is worth

Done properly, a plant typically holds fewer part numbers, more pieces of each, less total capital, and has the right part on the shelf when it matters. The last of those is the whole point. During a breakdown, being able to substitute from stores in ten minutes is worth more than an optimal specification for every individual position.

We will do the mapping for you

This is work we do regularly and it costs you nothing to ask. Send us your current sensor list, whatever state it is in, as a spreadsheet, a photograph of the stores shelf, or an export from your maintenance system. We will map it to the smallest set of equivalents that covers every position, flag anything over-specified, and quote the consolidated list.

We manufacture the D30C inductive range from M8 to M50 in Mumbai, alongside capacitive, photoelectric, magnetic and NAMUR switches, so one consolidated list can usually be covered from one supplier with one lead time.