Do You Know How to Improve Your Maintenance Quality?

Where post-service failure fits in Defect Elimination

Defect elimination starts from the observation that a large share of in-service failures aren't random bad luck — they trace back to a handful of avoidable causes: maintenance carried out poorly, mal-operation such as overloading, environmental stress like abrasive dust, or poor-quality inputs and outputs such as contaminated fuel. The goal is to find and remove those root causes, closing the gap between the reliability an asset was actually designed for and the reliability it delivers in the field.

Post-service failure analysis targets the first of those causes directly: maintenance-induced failure. The logic is straightforward. Any failure that shows up shortly after a maintenance intervention is far more likely to be a consequence of how that maintenance was done than a coincidence. Not every quality lapse causes an outright failure — a fastener left under-torqued, for instance, might only mean an asset comes back offline briefly for rework — but the pattern is the same underlying problem.

Making it measurable

To find post-service failures systematically, you need your CMMS to reset a component's age clock to zero every time it's touched by maintenance, so you can see how long it survived after the last intervention rather than since it was first installed. For fleets with more established reliability practices, Weibull analysis is a natural fit here: a component distribution with a shape parameter between 0 and 0.9 falls into the "premature failure" pattern — exactly the signature post-service defect elimination is looking for. Crow-AMSAA analysis, borrowed from the way manufacturers track reliability growth during new-product development, can add a second lens by flagging when a component's recent failure trend has quietly turned for the worse.

Weibull's B-5 age — the age at which 5% of a population is expected to have failed, read off the cumulative distribution — gives a workable rule of thumb: out of every 20 failures, only about one should reasonably occur before the B-5 age. Any failure younger than that is worth flagging for root cause analysis. Where Weibull data isn't available yet, the same idea can be approximated using experienced judgement from senior maintainers — a fixed threshold such as 100 operating hours, or a week of elapsed time since the last intervention, with anything failing inside that window treated as a candidate post-service failure.

Where maintenance work creates its own risk

It helps to think about which parts of a maintenance task are inherently more likely to introduce a defect:

• Breaking a seal to access internals increases the risk of dirt or moisture ingress afterward.
• Poor physical accessibility makes the work more awkward, and awkward work is more likely to be rushed or done imperfectly.
• Needing to remove other equipment just to reach the target component — sometimes called "work in wake" — adds extra steps where something can go wrong.
• Even slight misalignment when rotating machinery is reassembled can meaningfully shorten the life of a shaft or coupling.

The human and organisational angle

James Reason's work on organisational accidents offers a useful lens here too, even outside the safety-critical context it was originally developed for. His concept of "latent failures" captures exactly the kind of accessibility and design issues above, but it extends further — into whether procedures, training and experience are genuinely fit for the task, and into working conditions like fatigue that quietly increase the chance of a slip. General shop-floor quality disciplines matter as well: keeping workspaces clean and organised (the "Five S" approach) and using shadow boards so every tool is accounted for before equipment is closed back up are simple, low-cost ways to reduce the odds of a maintenance-induced defect in the first place.

None of this replaces good root cause analysis once a post-service failure is confirmed — but knowing where to look first, using a consistent age-based trigger tracked automatically through IronMan®, is what turns post-service failure from an occasional observation into a repeatable part of a maintenance quality programme.