Asset criticality score
Rank an asset before you spend PM dollars on it — worst-case consequence times failure likelihood, with a recommended strategy from run-to-failure up to full RCM. Not every machine deserves the same attention.
Asset criticality · consequence × likelihood
Criticality = worst consequence × likelihood (1–25)
What this gives you
Asset criticality is the number that tells you which machines earn your preventive-maintenance dollars and which do not. It combines how bad a failure would be with how likely that failure is: criticality = worst-case consequence × failure likelihood. Score each on a 1–5 scale and the product lands between 1 and 25, which sorts your equipment into four bands — LOW, MEDIUM, HIGH and CRITICAL — each with a maintenance strategy that fits the risk. It is the first step of any reliability-centered maintenance (RCM) program and the fastest way to stop treating every machine the same.
Why the worst consequence, not the sum
A failure has several kinds of fallout — someone could get hurt or a containment could be breached, production could stop, and the repair itself could be expensive. This calculator takes the maximum of the three consequence scores, not their average or their sum. The reason is simple: a machine whose failure could injure an operator is a high-consequence asset even if it is cheap to fix and barely dents production. Averaging would let a low repair cost dilute a safety hazard, which is exactly the mistake criticality analysis exists to prevent. Rate each consequence honestly, and the worst one drives the score.
Field note — likelihood is where optimism hides
The consequence scores are usually easy to agree on; the likelihood score is where a review goes soft. Teams anchor on "it hasn't failed yet" and rate a tired 15-year-old pump a 2. Use the failure history, the condition data and the age of the machine, not the memory of the last person to touch it. If you have no data, a machine past its design life with no redundancy is a 4 or a 5, not a hopeful 2 — and moving it one notch can jump the score across a band boundary and change the whole maintenance strategy.
The bands and what to do about them
A score of 5 or less is LOW — run it to failure and keep a spare. 6 to 12 is MEDIUM — a time-based PM on a calendar or runtime interval is the right level of effort. 13 to 19 is HIGH — move to condition-based predictive maintenance (vibration, thermography, oil analysis) so you catch the failure developing. 20 to 25 is CRITICAL — the machine warrants redundancy or a full RCM analysis; you do not want its failure to be a surprise or a single point of loss.
Worked example
A transfer pump: a failure could cause a minor environmental release (safety/environment 4), would halt one line but not the plant (production 3), and is a moderate repair (cost 2). The worst consequence is 4. The pump is aging and has failed twice in three years, so likelihood is 4. Criticality is 4 × 4 = 16, which lands in the HIGH band. The recommendation is condition-based PdM — put it on a vibration route rather than waiting for it to seize.