Obsolescence management: the stages, the ways out, what sets the risk and the register that holds it
What obsolescence is and why it is certain for electronics and software, the supplier's life cycle stages and the plant's move at each, the seven ways out and what each costs, what actually sets the risk rather than age, the register that holds every item at risk with its years of cover, and designing it out. With a free register template.
Key takeaways
- Obsolescence is a supplier ending the production or support of something a plant depends on, before the plant has finished depending on it. It is certain for electronics and software, likely for drives and instruments, and possible for anything made by a company that can be bought.
- A supplier's product passes through stages, from active through last time buy and end of sale to end of support, and the plant's choices get dearer at each one. The cheap choices are made before the last time buy.
- The ways out are a last time buy, an alternative source, a repair route, the successor product, a bridge that presents the old interface, salvage, or replacement. Each has a cost and a catch, and the register says which was chosen and why.
- The risk is set by criticality, support status, spares held against usage, lead time, failure rate, skills and software, not by age. A thirty year old gearbox with a maker still in business is less of a problem than a six year old controller whose series was discontinued.
- The register is the instrument: every item at risk, its status, its cover, its resolution and its owner, reviewed on a date. An obsolescence found by a failure, with no spare and no source, is a register that was not kept.
A processor module fails on a Sunday. The one spare went in two years ago. The supplier discontinued the series in 2024 and the last units on the resellers' shelves went to someone else. The plant runs for three days on a module bought from a broker at eleven times the original price, with no warranty, and the engineering manager finds out that the controller replacement project that would have prevented all of it was proposed in 2023 and deferred.
That is obsolescence found by a failure, and it is the expensive way to find it. The cheap way is a register: every item the plant depends on that a supplier could stop making or supporting, its status on the supplier's life cycle, the spares held against the rate they are used, the resolution chosen and the date by which it has to be done. The choices are cheap before the last time buy and dear after end of support, and the register is what makes the cheap choice possible.
This guide sets out what obsolescence is and why it is certain for some things and merely possible for others, the stages a supplier's product passes through and what the plant does at each, the ways out and what each costs, what actually sets the risk, the register that holds it all, and where obsolescence management fails. A register template comes with it. It sits beside the critical spares guide, which covers the holding that buys time, and the asset lifecycle guide, which covers the replacement that ends the problem.
Free download
Obsolescence register template
An obsolescence register with the item, where it is used, its criticality, the supplier's life cycle status and dates, spares held against annual usage with the years of cover worked out, the resolution, the plan and the owner, five worked rows across a controller, a drive, an operating system, an instrument and a seal, and the stages and the resolutions on sheets of their own.
What obsolescence is
Obsolescence is a mismatch of lives. A plant is built to run for twenty or thirty years. The things it is built from have product lives set by their makers, and for electronics and software those lives are a fraction of the plant's. A controller series is sold for a decade and supported for a few years after. An operating system has an end of support date published years ahead. A drive model is superseded by one with a different frame and different control signals. None of that is a defect. It is how the suppliers' industries work, and a plant that depends on their products inherits their calendar.
For mechanical items the problem is different and slower: a maker is bought, a product line is dropped, a foundry closes, a pattern is lost. The item can often be made again by someone, at a cost and a lead time, which is why the mechanical rows on a register are mostly about identifying a second source rather than racing a published date. The register has both kinds, and the first kind is sorted to the top.
The stages, and the plant's move at each
Suppliers of control, drive and electrical products run their product lines through published stages, and the plant's choices get dearer at each. While an item is active the plant records it and watches. When it is mature, with a successor announced, the plant checks whether the successor is a drop-in and plans the migration on its own timetable. The last time buy notice is the point of decision: a lifetime buy, an alternative, or a funded replacement, chosen now with the money, because after end of sale the plant buys what it can find and after end of support it runs what it has.
The table under the figure gives the plant's move at each stage. The supplier's status is a column on the register and the end of sale and end of support dates are two more, and keeping them current is the main work of the register: a check of each critical supplier's life cycle notices, quarterly, by the engineer who owns that category. Suppliers publish these for control products and most drives. For instruments and mechanical items the status is often unknown until asked, and the register says unknown rather than guessing.
1Active
In production, supported, the current product
The plant
Record it on the register with its criticality, and the supplier's published life cycle status if there is one
2Mature
Still sold and supported, but a successor exists and the supplier's effort is moving to it
The plant
Note the successor, check whether it is a drop-in, and plan the migration before it is forced
3Last time buy
The supplier announces the end of production and a date by which final orders are taken
The plant
Decide the strategy now: the lifetime buy, the alternative, or the replacement project, with the money
4End of sale
No longer sold new. Repair and support continue for a published period
The plant
Hold the spares the strategy says, set up the repair route, and schedule the replacement
5End of support
No repair, no firmware, no security patches, no engineering help
The plant
The replacement is due. Anything still running is on the plant's own risk, written down
6Unsupported
Running on salvage, the grey market and the people who remember it
The plant
A critical item here is a finding in the register and a line in the risk register
Choices are cheap up to here
→ From here the plant pays whatever the market asks
The ways out
There are seven resolutions, in rough order of cost, and the table gives each with its catch. A last time buy is cheap and easy and depends on a forecast of usage over the remaining life, with a margin, and on the parts surviving on the shelf, which electronics do not always do. An alternative source needs an engineering check that the equivalent is one: for a bearing that is a datasheet comparison, for a controller module it is a firmware, protocol and software tool comparison that often fails. Repair needs a route with a turnaround the holding can cover. The successor product is the supplier's own answer and is a drop-in less often than the brochure says. A bridge, a protocol gateway or a drive that accepts the old control signals, buys time for interfaces and nothing for mechanical parts. Salvage from decommissioned equipment is a stopgap. Replacement is the only permanent answer and the only expensive one.
Most items end up with two resolutions: one that buys time, and the replacement that ends the problem, with a date. The register holds both, and the critical spares holding is where the time-buying resolution lives.
| Resolution | What it is | The catch |
|---|---|---|
| Last time buy | Buy enough before the end of production to last to the planned replacement, plus a margin for the failure rate | Ties up money; the parts age on the shelf; the forecast has to be right |
| Alternative source | A second maker's equivalent, a form-fit-function replacement, or an authorised reseller's stock | Needs an engineering check that the equivalent is one; software and protocols often are not |
| Repair | A repair route for failed units: the maker, a specialist, or the site's own workshop with the drawings | Turnaround time has to fit the spares holding; some items cannot be repaired to the original standard |
| Successor product | The supplier's replacement, where it is a drop-in or close to it | Firmware, mounting, wiring and software tools change more often than the datasheet admits |
| Emulation or bridge | A new device that presents the old interface: a protocol gateway, a drive that accepts the old control signals | A second thing to maintain; works for interfaces, not for mechanical parts |
| Redesign or replace | A project to replace the equipment or the system, planned to a date | The expensive answer, and the only permanent one |
| Salvage | Parts from decommissioned equipment, held and tested | A stopgap with no supply behind it |
Buy time with the cheap resolution, and spend the time on the replacement. The time runs out either way.
What sets the risk
Age is the factor people reach for and it is the wrong one. A thirty year old gearbox from a maker still in business with the drawings on file is a low risk. A six year old controller whose series was discontinued last year, with one spare, in the crusher's control panel, is a high one. The table gives what actually sets the risk, and the register carries each as a column so that the items can be sorted by it. The years of cover, which is the spares held divided by the annual usage, is the single most useful number on the register, and the template works it out.
Two factors are routinely missed. Skills and tools: an item that nobody on site can configure, whose programming software runs only on an operating system the IT department has retired, is obsolete in practice whatever the supplier says. And software and licences: an HMI on an operating system past its end of support is an OT security finding as well as an obsolescence one, and a licence that cannot be transferred to new hardware turns a hardware failure into a software replacement.
| Factor | What it means for the risk |
|---|---|
| Criticality of what it is in | An obsolete card in a critical system is a different problem from the same card in a standby one |
| Support status | Where the item is on the supplier's life cycle, and whether the supplier publishes it at all |
| Spares held against usage | Years of cover at the current failure rate, counting the spares that will fail on the shelf |
| Lead time and source | Whether a replacement can be bought at all, from whom, and how long it takes |
| Failure rate and age | An item failing more often as it ages consumes the holding faster than the plan assumed |
| Skills and tools | Whether anyone can still configure it, and whether the software and the laptop that runs it still exist |
| Software and licences | An operating system out of support, a licence that cannot be transferred, a vendor that no longer issues keys |
| Interfaces | What else has to change if this does: the protocol, the wiring, the cabinet, the HMI |
The register
The register is one row per item at risk, with the columns in the table. The item is tied to the equipment it is used in by tag, so that the register can be read from the asset: a criticality assessment on the crusher finds the controller, the drive and the seal on the same row set. The supplier's status and dates come from the supplier's notices. The spares and usage come from the stores system and the work order history, and the years of cover from dividing them. The resolution, the plan and the owner are the decision, and the review date is what brings the row back.
The template has five worked rows: a discontinued processor with a funded replacement, a drive past end of support with its successor in progress, an operating system with a published end of support date, an instrument resolved by an approved alternative, and a seal from a maker that was bought out, status unknown. They are the five kinds of row most registers contain. At a larger site the register is held in the CMMS or the engineering document system against the equipment records. The columns are the same.
| Column | What goes in it |
|---|---|
| Item | The part or product, as the supplier names it |
| Category | As it says |
| Make and model | As it says |
| Where used | The equipment tags, so the register can be read from the asset |
| Quantity installed | As it says |
| Criticality | Of the equipment it is in |
| Supplier status | As it says |
| End of sale | The supplier's date, or blank |
| End of support | As it says |
| Spares held | As it says |
| Annual usage | Failures per year, from the history |
| Years of cover | Spares held divided by annual usage, worked out on the sheet |
| Resolution | As it says |
| Plan | What will be done, by when, and the project or work order it sits in |
| Owner | As it says |
| Review date | As it says |
| Status | As it says |
Designing it out
IEC 62402 puts a good part of obsolescence management at the design stage, where it is cheapest, and the design brief and the operational readiness review are where a plant can ask. Standardise on the fewest product families so the holding is smaller and the migration is one project rather than ten. Prefer open protocols and published interfaces, which a bridge can present later. Ask the supplier for the product's life cycle stage and the planned end of sale before buying, not after. And write the register at handover, from the asset register, rather than discovering the first discontinued module in the field.
The management of change process is the other place obsolescence enters. A replacement in kind is a replacement with the same item, and an obsolete item cannot be replaced in kind. The change that fits the successor is a change, with the review that implies, and the register is where the change is planned rather than forced.
Where it fails
- No register, so obsolescence is discovered by the failure that makes it urgent.
- A register of the mechanical spares only, with the controllers, drives, networks and software, which are where the published dates are, not on it.
- Risk judged by age, so the old gearbox is watched and the young controller is not.
- A last time buy sized without a usage forecast, or stored without testing, so the lifetime stock fails on the shelf.
- The successor product assumed to be a drop-in, and the cubicle modification, the software tool and the retraining found at the stop.
- Supplier life cycle notices sent to a mailbox nobody reads.
- The replacement project proposed, deferred, and proposed again after the failure.
- An operating system past end of support treated as an IT matter, with the control system engineer not told.
Common questions
What is obsolescence management?
The practice of knowing which of the items a plant depends on are at risk of no longer being obtainable or supported, deciding in advance what will be done about each, and doing it before the failure that makes the decision urgent. IEC 62402 sets out the requirements: a policy, a plan, a register of items at risk with their status, the resolutions chosen, and the tracking of how the plan is performing. On a plant it is owned by engineering, with inputs from maintenance, stores and the control system owner.
What is a last time buy?
An order placed before a supplier ends production, for enough of the item to last to its planned replacement plus a margin for the failure rate. Suppliers announce a last time buy date, after which the item cannot be bought new. It is the cheapest resolution when the replacement is some years away and the item is small, and the worst when the forecast is wrong: too few and the plant is on the grey market, too many and the parts age on the shelf. Electronics in particular degrade in storage, and a lifetime buy is tested periodically.
What is the difference between end of sale and end of support?
End of sale is the date after which the supplier no longer sells the item new. Repair, spares and engineering support usually continue for a published period after it. End of support is the date after which they stop: no repair, no firmware, no security patches, no help. Between the two a plant can still keep an item running through the supplier. After end of support it is on its own, and for a control system or software component that is also a security finding.
Why is obsolescence a bigger problem for control systems?
Because electronics and software have short product lives and long plant lives. A processor series is sold for ten years and supported for five more, and the plant it controls runs for thirty. The operating system under the HMI reaches end of support on a published date whether or not the plant is ready. And a control system's parts are not interchangeable across generations, so one obsolete module can force a replacement of the controller, the IO, the software tools and the engineer's laptop together. That is why the register starts with the controllers, the drives, the networks and the software.
How do you assess obsolescence risk?
Per item, from what decides the consequence and the likelihood: the criticality of the equipment it is in, the supplier's published life cycle status, the spares held against the annual usage as years of cover, the lead time and whether a source exists at all, the failure rate and whether it is rising with age, whether the skills and the software tools to configure it still exist, and what else would have to change if it did. Age on its own is not a factor. The register carries each of those, and the years of cover column is the one to sort by.
What goes in an obsolescence register?
Every item at risk: the part or product, its category, make and model, the equipment it is used in by tag, the quantity installed, the criticality, the supplier's status and the end of sale and end of support dates, the spares held, the annual usage, the years of cover that follow, the resolution chosen, the plan with its date and the project or work order it sits in, an owner, a review date and a status. The template with this guide has those columns and five worked rows.
Key terms
Plain-language definitions from our glossary for the concepts this article leans on.
Standards and further reading
- IEC 62402:2019 Obsolescence management (IEC)
- ISO 55001:2024 Asset management system requirements (ISO)
- IEC 60300-3-3:2017 Dependability management, application guide: life cycle costing (IEC)
- ISO 14224:2016 Collection and exchange of reliability and maintenance data for equipment (ISO)
- Principles of operational technology cyber security (Australian Signals Directorate)
Related case studies and tools
- Life Cycle Cost Calculator (tool)
- Asset management planning and the system behind it (case study)
Related reading
Working through something like this?
See how we approach these initiatives, or tell us what you are dealing with.