Asset health scores, and when they are worth having

An asset health score condenses age, condition, failure history and cost into one number per piece of equipment, so a portfolio can be sorted. It is genuinely useful for capital planning and routinely useless for maintenance decisions, and knowing which is which saves a lot of wasted effort.

What belongs in the score

• •Age against expected life, which you have for free
• •Condition from the last inspection, scored on a fixed scale
• •Failures in the last twenty-four months
• •Repair spend against replacement cost
• •Parts availability, because an asset you cannot get parts for is unhealthy whatever its condition

What to leave out

Anything you cannot get consistently for every asset. A score that uses vibration data for the ten machines that have sensors and guesses for the other four hundred is not a portfolio view, and it will quietly rank the instrumented machines worse because you know more about them.

Keep the inputs visible

Never show the score alone. Show the four or five components next to it. A score of 42 tells nobody what to do. "Age 18 of 20 years, condition poor, 9 failures, repairs at 60 percent of replacement cost" is a replacement case that writes itself.

Where it is genuinely good

Capital planning across hundreds or thousands of assets, where somebody has to choose what to replace next year and cannot look at each one. That is a real problem and a health score is a reasonable answer to it.

Where it is not

Deciding what to do today about one machine. For that, look at the machine. A score is a summary of things you already know, and summarising them loses exactly the detail a decision needs.

Do not let it drift

A health score is only as fresh as its condition data. If inspections are annual, the score is annual, and presenting it on a live dashboard implies a currency it does not have. Put the date of the underlying condition assessment next to every score, always.