SiteThe filler on line 1 has stopped for the third time this month. Same fault, same drive, same trip. You'll know the fault that keeps coming back. Production is fed up. The engineering manager calls a meeting for ten o'clock.
Everyone turns up with their story ready. The operator says it was running fine. The day fitter says the night fitter didn't fix it properly. The night fitter isn't there, because they're asleep. Twenty minutes later the meeting ends with "keep an eye on it", and the room is a bit colder than it was.
That's how most root cause meetings go wrong. They turn into a hunt for who did it. Once people feel hunted, they stop telling you what happened.
Five whys is a simple tool. Used well, it gets past the first easy answer to the cause that will keep stopping the line. Used badly, it's just a blame session with a whiteboard.
How it works
Start with the event. Ask why it happened. Write the answer down. Ask why again, of that answer. Keep going until the answer stops being useful.
HSE uses the same method in its guide to investigating accidents. It describes "the simple technique of asking 'why' over and over, until the answer is no longer meaningful" (HSG245). The guide is written for accidents, but the thinking works just as well on a breakdown.
Don't get hung up on the number five. HSG245 says not to worry about how many times you ask, because doing so is how you reach the real causes. Sometimes it's three. Sometimes it's seven.
What it looks like on a real fault
Take the filler.
- Why did the filler stop? The drive tripped on overcurrent.
- Why did the drive trip? The motor was drawing too much current at start-up.
- Why? The gearbox is stiff to turn. The bearing on the output shaft is rough.
- Why is the bearing rough? It's running dry. There's hardly any grease in it.
- Why is it dry? It isn't on the PPM sheet. The grease point sits behind a guard that was added two years ago, and the sheet was never updated.
Look at where that ends. Not with the night fitter. Not with the operator. With a PPM sheet that didn't change when the machine did. Fix that, and you fix the trip. Reset the drive, and you'll be back in the same meeting next month.
Keep the whys on the machine and the system
Here's our position. Every "why" should be about the machine, the method or the system. If the answer to a why is a person's name, ask a different question.
So "Why was it dry? Because the fitter didn't grease it" becomes "Why wasn't it greased? Because it wasn't on the sheet." The first answer ends the conversation. The second one leads somewhere.
HSE is clear on this. HSG245 calls blaming individuals "fruitless". It also warns that trying to "apportion blame before the investigation has started" is counterproductive, "because people become defensive and uncooperative" (HSG245).
It goes further on what happens next. Unless you find a deliberate and malicious act, HSE says disciplinary action "may be counter-productive", and asks a fair question: "will anyone be open and honest with you the next time" (HSG245). On a shop floor, the next time is usually next week.
None of this means nobody is ever responsible. HSG245 talks about a fair and just system, where people answer for their behaviour without being unduly blamed. But a five whys on a breakdown is for finding the fault, and people stop helping as soon as it feels like a trial.
Where root causes usually end up
If you keep asking, most chains end in the same few places. HSG245 says root causes "are generally management, planning or organisational failings" (HSG245). On a maintenance team, that tends to mean one of these:
- a PPM task that's missing, out of date or at the wrong interval
- a spare that wasn't held, so a worn part was left in
- a skill nobody on that shift had
- a modification nobody wrote down
- a handover that didn't happen
- an operator who was never shown the early signs
The skills one comes up more than people expect. A drive fault waits four hours on nights because nobody on shift is confident with drives. The why chain ends at the rota, and the fitter was never the cause. You have a cover problem, and it has a cost you can put a number on.
Run it in twenty minutes
You don't need a long meeting. You need the right people, and the facts in front of them.
Do it soon. The same day if you can, while memories are fresh and the old parts are still on the bench.
Get the people who were there. The operator, the fitter who fixed it, and the shift manager. If the night fitter fixed it, hold it at shift change or ring them in. Don't hold it without them and then decide what they did.
Bring the evidence. The breakdown log entry, the HMI photos, the fault history, the parts that came out. Our post on a breakdown log that actually tells you something covers what you need to have written down.
Write the chain where everyone can see it. A whiteboard by the machine works. Ask each why out loud and write the answer before anyone moves on.
Agree one action, with a name and a date. Not a list of ten. The one change at the bottom of the chain that would have stopped this. Then check it was done. Our post on the breakdown debrief covers making this a habit after every big stop.
If the chain splits, follow each branch. A fault often has more than one cause, and the guide says some lines of enquiry will end sooner than others. For a fault that's been back four times, you may need a fuller method. Our post on root cause methods for maintenance teams covers the options.
Keep production in the room, but not in charge
Production should be there. They saw it stop and they live with it. But a five whys run by whoever is angriest soon becomes a row. Our post on production versus maintenance at 3am goes into that.
The person running it asks the questions. They don't answer them, and they don't let anyone answer with a name.
What to do on Monday
- Pick the fault that has come back most this month and run a five whys on it this week.
- Agree the rule with your team before you start: no names as answers.
- End with one action, a named owner and a date, and check it at the next weekly review.
- Look back over your last few root cause meetings. If they ended at a person, run them again.