Criticality ranking: deciding which machines matter most

Ask the production manager which machines are critical. You'll get a list of every machine on site.

Ask the night shift fitter and you'll get a different answer. The one with no spare drive. The wrapper that only one person can set up. The compressor that takes the whole site down when it trips.

Both answers matter. But if everything is critical, nothing is. Your PPM hours, your spares budget and your cover all get spread thin across kit that doesn't need them. Meanwhile the one machine that can stop the site gets the same attention as the bin lifter.

A criticality ranking fixes that. It's a list of your machines in order of how much it hurts when each one stops. Nothing more clever than that.

Why bother

You already make these calls every day. When two machines go down at once, someone decides which gets the fitter. A ranking just makes that call once, on a quiet afternoon, rather than at 3am with two shift managers shouting.

It also shapes the rest of your planning. The PPM schedule, the spares you hold, and who you need on which shift should all follow from it. If you're trying to get the balance between planned and breakdown work right, this is where it starts.

And there's a safety reason. HSE's code of practice for PUWER says the frequency of maintenance should take account of the "risk to health and safety from malfunction or failure" (L22, paragraph 72). It goes on to say maintenance should be "targeted at the parts of work equipment where failure or deterioration could lead to health and safety risks" (L22, paragraph 73). So a ranking that ignores safety misses the point.

Keep the scoring simple

You don't need software or a consultant. A spreadsheet and a couple of hours with the right people will do. Get your best fitter, a shift manager and someone who knows the order book in the same room.

Score each machine from 1 to 3 on five questions. These are the ones we'd use:

1. Safety: if it fails, could someone get hurt? Think guards, interlocks, pressure, stored energy and anything that moves.
2. Output: does the whole line or site stop, or just one lane? Is there another machine that can pick up the work?
3. Cost of a stop: lost output, idle labour, overtime to catch up, scrap on restart and late delivery penalties.
4. Time to fix: how long is a typical repair, and how long would the parts take on a bad day?
5. Who can fix it: can anyone on shift deal with it, or only one person?

Add the scores up. The machines at the top of the list are the ones that matter most. Don't spend long arguing over a single point. The order matters more than the exact score.

For the cost question, put real numbers in if you have them. Our guide on working out what an hour of downtime costs shows how. And if you don't have a list of what's on site to start from, build that first. Asset registers covers the basics.

The question most rankings miss

Most criticality scores stop at safety, output and cost. We'd always add the fifth one: who can fix it.

Picture two machines. Both stop the line. Both cost the same per hour. One is a simple conveyor any fitter can sort in half an hour. The other is a robot cell with a servo system only one of your engineers really understands.

On paper they score the same. In real life, the robot cell is far more dangerous to lose. When it fails on a Saturday night and that one engineer is off, you're looking at a long stop. You're paying for every hour of it while you hunt for someone who knows the kit.

So score the people as well as the machine. If a machine needs a skill that only one person on site has, push it up the list. Then do something about it. Train a second person. Get the drawings and backups in order. Line up an outside engineer who knows that make. Our guide to single-skill risk goes into the options.

Don't forget the hidden faults

Production will always shout about a machine that's stopped. They won't shout about a safety device that has quietly failed.

L22 makes the point well. A fault that affects production "is normally apparent within a short time". A fault in a safety-critical system "could remain undetected unless appropriate safety checks are included in maintenance activities" (L22, paragraph 71).

So a light curtain, an interlock or a pressure relief valve can score high on safety even when it scores low on output. Give those parts their own checks on the PPM, whatever the machine they sit on scores overall.

Put the ranking to work

A ranking on a shared drive does nothing. Use it.

• Your top machines get the most planned work, and their PPM never gets cancelled for a breakdown elsewhere.
• Their spares get held, with lead times checked. See critical spares.
• Their manuals, drawings and PLC backups get found and stored where the night shift can reach them.
• When two machines are down at once, the higher one gets the fitter first. No debate.
• Your cover plan makes sure someone who can fix the top machines is on every shift, or on the phone.

Look at the bottom of the list too. Some of those machines are fine to run to failure, as long as there's a spare and someone can fit it.

Review it once a year, and whenever a line changes. A new machine, a new customer or a lost backup can move kit up the list overnight.

What to do on Monday

1. Book two hours with your best fitter, a shift manager and someone from planning. Score your machines on the five questions.
2. Take the top five and check each one has its PPM, its spares and its backups in place.
3. For each of the top five, write down who on site can fix it. If it's one name, plan for a second.
4. Share the list with every shift manager, so the 3am call is already made.

* Paragraph numbers refer to L22, fourth edition (2014, amended 2018), correct as of 8 September 2026. Check HSE's L22 page for the latest.