Finding the equipment that is eating your year

In almost every estate, a small number of assets cause a large share of the unplanned work. Finding them takes an hour with data you already have, and it is the highest-return hour available to most maintenance teams.

How to find them

Export last year’s unplanned work, group by asset, and sort by count. Then do it again sorted by hours, and again by cost. The three lists overlap but not completely, and each one is telling you something different.

• •By count: what is disrupting the team most often
• •By hours: what is consuming the capacity
• •By cost: what is consuming the budget, which is usually a shorter and more expensive list

If the work is not attached to assets, start there

Most teams cannot run this analysis because jobs are filed against buildings rather than equipment. That is fixable from today: a code on the machine that opens the job for that machine. In twelve months you will have the data. Nobody has ever regretted starting that clock.

Then ask why, once, for each one

For each asset on the list, read the closing notes on every job. This is why the closing note matters more than the cause code: sixty jobs saying "attended, reset" tell you nothing, and twenty saying "tripped on high pressure again" tell you exactly where to look.

The four answers you will find

• •The preventive work is not addressing the failure mode that keeps happening
• •The equipment is being operated outside what it was specified for
• •It was installed badly and has never worked properly since
• •It is at the end of its life and is being repaired past the point where that makes sense

Only the first is fixed by changing the maintenance schedule. The other three need somebody outside maintenance to agree to something, which is why this analysis needs to be presented rather than filed.

How to present it

One page, five assets, with the count, the hours, the cost and one sentence on the cause. Then the ask. A replacement request supported by "this unit has failed 23 times in 14 months and cost 41,000 in repairs" is a different conversation from a general request for capital.

Then check it worked

Whatever you change, write the date on it and look again in six months. Half of bad actor fixes do not work, and the ones that do are the evidence that funds the next round.