When the failed component has vanished from the catalogues, the repair becomes a matter of understanding: what function did the part serve in the machine, and which equivalent can fill that role. The blockage rarely comes from procurement, it comes from knowledge of the machine.
On an asset base approaching twenty years old, the situation repeats itself: a PLC at the end of manufacturer support, a variable speed drive that is no longer produced, a sensor replaced by a generation with different characteristics. Each time, the team starts the search again. Industrial intelligence platforms such as Mimorian model the equipment and structure fault diagnosis, so that this reasoning, once carried out, stays available for the jobs that follow.
Why does an obsolete part cost more than its price?
Ordering an equivalent component takes a few clicks, once you know which one to order. All the time is spent before that.
The technician first has to establish what the original part did. On an ageing machine, that information lives in three places that contradict each other: the manufacturer's drawing, often predating the successive modifications, the actual machine as it was rebuilt through repair after repair, and the memory of whoever was there during those repairs.
Then comes the question of compatibility, which goes well beyond dimensions. A sensor with the same form factor may deliver a different signal. A drive of equivalent power may need settings that the old one received by default. These gaps show up at restart, when the line has already been put back together.
The real cost therefore sits in the hours of searching and in the back and forth, rather than on the supplier's invoice. This waste has a known cause: inefficient knowledge sharing costs a large business an average of 47 million dollars a year, in lost productivity and rework [Source: Panopto, 2018]. On an ageing machine, that cost takes the very concrete form of a search a colleague already carried out three years earlier.
Which questions come before looking for an equivalent?
The replacement is prepared by first establishing what the part does, before looking for what resembles it. Four questions cover the majority of cases.
| Question | What to establish | Where to find the answer |
|---|---|---|
| Which exact function? | The role of the part in the chain, not its reference | Original drawing, survey on the machine, reasoning of a long-serving colleague |
| Which interfaces? | Signal, power supply, protocol, mechanical mounting | Manufacturer documentation, direct measurement on the installation |
| Which behaviour on fault? | What the machine does when this part fails | Fault history, job reports |
| Which regulatory constraints? | Machine safety, site requirements, certifications | Technical file, methods department |
The third row often gets neglected, and it is the one that gets paid for at restart. A fallback component that fills the nominal function but behaves differently on fault turns a clean stop into an incident.
The fourth deserves the same attention on machines that have seen several generations of standards.
How does today's replacement serve tomorrow's failure?
An equivalent found and validated represents several hours of reasoning. Those hours stay available or disappear depending on what gets written after the job.
A useful job report holds four elements: the original part and its function, the equivalent chosen, the gaps found during fitting or setting up, and the behaviour observed at restart. Recorded in a queryable field, these elements make the second occurrence far faster than the first. Recorded in a free-text field, they become impossible to find.
This memory carries all the more weight as the people who know are becoming scarce: 89 % of American industrial executives recognise a talent shortage, and 2.4 million posts could stay unfilled by 2028 [Source: Deloitte/Manufacturing Institute, 2018]. On an ageing asset base, whoever knows the history of the modifications is often the one approaching retirement.
The gap between the original drawing and the actual machine grows with every non-standard replacement. Documenting the gap at the moment it is created keeps the representation of the machine up to date. We cover this approach in our article on the functional digital twin in maintenance, and on industrial memory that gets lost.
Frequently asked questions
Should the machine be replaced when parts become impossible to find?
Replacement is justified on criticality and on the cumulative cost of downtime, rather than on the availability of components alone. Many ageing machines stay reliable once the question of equivalents is handled methodically. The comparison rests on the cost of the stops suffered, which we cover in our article on the real cost of unplanned downtime.
How do you know whether an equivalent component will do the job?
By starting from the function it fills rather than from the reference. The four questions in the table above cover the function, the interfaces, the behaviour on fault and the regulatory constraints. An equivalent that passes those four points is then validated at restart.
What should be done when the drawing no longer matches the machine?
Survey the actual installation on the part concerned and record the gap. A partial but accurate survey serves better than a complete drawing that has become false. The gap documented at the time of replacement avoids redoing the survey at the next failure.
Who should keep the record of validated equivalents?
The maintenance team, in a support that can be queried by equipment. A personal spreadsheet disappears with its author. The record is worth what it can bring back at the moment of the failure, to a technician who was not there the first time.
Conclusion
Three points to remember.
- The blockage is about knowledge, rather than about procurement. Once the function is established, the equivalent can be found.
- The method holds in four questions. Function, interfaces, behaviour on fault, regulatory constraints: that order avoids the bad surprise at restart.
- The record is worth as much as the part. What gets written after the job decides how much time the next occurrence will cost.
Next step: identify the equipment in your asset base whose critical component is at the end of manufacturer support, and check whether an equivalent has already been validated somewhere in your job reports.