> ## Content Index
> Fetch the complete content index at: https://the-readiness-directive.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# When Automation Outlives Its Cyber Support
- URL: https://the-readiness-directive.ghost.io/untitlewhen-automation-outlives-its-cyber-supportd/
- Published: 2026-11-02T07:30:00.000Z
- Updated: 2026-11-02T08:19:22.000Z
- Description: Why cyber end-of-support creates a different lifecycle problem for safety-related control systems.
- Author: William Corkhill
- Tags: Industry Spotlight, Industrial Automation, Robotics, Machinery Regulation, cybersecurity, Product Integrity

### *Why cyber end-of-support creates a different lifecycle problem for safety-related control systems.*

**Industry Spotlight — Industrial Automation & Robotics**

In April 2025, CISA published an advisory covering Schneider Electric Modicon controllers, including Modicon Quantum Safety. The vulnerabilities were not new: the CVE identifiers in the advisory date from 2018 and 2019, and CISA pointed users back to Schneider Electric’s 2019 security advisory for the detailed remediation history. The highest-rated vulnerability scored 10.0\. CISA reported no known public exploitation specifically targeting the vulnerabilities. \[1\]

What makes the case useful today is not the age of the vulnerabilities. It is what happened to the product around them.

CISA records that Quantum and Quantum Safety had reached end of life and were no longer commercially available. For the vulnerabilities listed against Quantum Safety, the advisory provides no firmware correction. Schneider Electric instead says users should strongly consider migration towards its current M580 or M580 Safety platform. At the same time, CISA lists compensating measures including application passwords, segmentation, access-control lists and restrictions around Port 502\. \[1\]

An unsupported controller does not automatically become unusable because a vulnerability exists. Compensating controls may reduce exposure and justify continued operation for a period. What those controls cannot do is restore the vendor-supported security path.

An industrial system can therefore reach an unusual lifecycle position: the controller still operates, spare hardware may still exist and the machine around it may have years of useful life remaining, while the supported means of correcting future vulnerabilities has ended.

The problem is familiar to OT cybersecurity practitioners. End-of-support has long been recognised as an obsolescence risk.

The more interesting question is what happens when that risk sits inside a **safety-related control system** and begins to drive machinery lifecycle decisions.

## Obsolescence already has a playbook

Industrial organisations have managed obsolete components for decades.

IEC 62402 formalises obsolescence management across the lifecycle, including design strategies, resolution selection and planning for systems that remain useful after individual technologies disappear. \[2\]

The practical responses are well established. Organisations make last-time buys, maintain spares pools, source reconditioned components, preserve application backups and plan successor-platform migrations into shutdown windows.

Machinery and functional-safety disciplines already expect safety-related modifications to be assessed and verified appropriately. IEC TS 63074 also deals directly with the relationship between cybersecurity threats and safety-related control systems where a security vulnerability could affect the ability to maintain safe machine operation. \[3\]

ISO/TR 22100-4 similarly addresses IT-security threats that can influence machinery safety. \[4\]

There is therefore little value in arguing that cybersecurity and machinery safety have suddenly discovered each other.

The interesting difference appears when the conventional obsolescence playbook stops solving the underlying exposure.

A spare controller can replace failed hardware. A reconditioned module can extend availability. A last-time buy can delay migration.

**A spare controller can preserve availability. It cannot restore security support.**

That is the mechanism, not the discovery.

The machinery question begins with what follows from it.

## The decisive date may not be the end-of-life date

Traditional obsolescence programmes often work around dates supplied by the vendor: end of sale, end of service, final support, spare availability.

Cybersecurity introduces another threshold.

A product can remain mechanically and functionally acceptable after commercial support has ended. Equally, a later vulnerability disclosure can change the risk associated with continuing to operate an unsupported platform even though nothing has physically failed.

Quantum Safety does not itself demonstrate that second scenario. The vulnerabilities in the CISA advisory were known before the platform reached the end of its active commercial life.

What it does demonstrate is that **known, unresolved vulnerabilities can survive beyond the commercial life of a safety platform**. \[1\]

The wider lifecycle risk is that a future disclosure may arrive only after security support has already ended.

For leadership, the more useful date may therefore be neither the end-of-sale date nor the day the last spare disappears.

It may be the point at which the organisation can no longer demonstrate that its combination of vendor support, compensating controls and planned migration remains an acceptable basis for continued operation.

That is a **defensibility threshold**.

It can arrive before the asset owner intended to replace the equipment.

## Quantum Safety shows where cyber and functional safety meet

Quantum Safety is predominantly a process-safety platform, so the CISA case should not be presented as though it were a machinery accident or a pure machinery application.

The underlying mechanism, however, is directly relevant to machinery using safety-related control systems.

When migration from an ageing safety controller becomes necessary, the task need not amount to redesigning the machine’s entire safety concept. A well-supported successor platform may have migration tools, established engineering procedures and known compatibility routes.

Even a straightforward migration still has a safety dimension.

New component data may require the achieved PL or SIL to be reverified, followed by validation of the affected safety functions. Safety I/O, diagnostics, communications or interfaces may also need assessment depending on the architecture.

That is routine functional-safety engineering.

The less routine part can be the reason and timing of the work. Cybersecurity can become the trigger for a migration that operations had expected to postpone for several more years.

## The operator may never have held the original design basis

This is where the problem becomes more than obsolescence management.

The operator owns the installed decision. It has to decide whether to continue operating, strengthen compensating controls, schedule a migration or replace the asset.

But the operator may never have held the original **design basis**: the risk assessment, safety requirements and engineering rationale behind the affected safety functions.

Under Regulation (EU) 2023/1230, manufacturers must retain specified technical documentation and the EU declaration of conformity for at least ten years after machinery or a related product is placed on the market or put into service. The Regulation also provides for relevant source code or programming logic to be supplied to competent authorities where necessary to demonstrate conformity. \[5\]

That technical documentation is principally part of the manufacturer’s conformity evidence. It is not the same thing as the operating documentation normally held by the owner of the machine.

On a machine operating for twenty years, the practical consequence can be significant.

The operator may have instructions, maintenance history and local modification records while lacking the original assumptions behind a safety function. The machine builder may no longer support the product. The original integrator may have disappeared. Ownership may have changed several times.

A later controller migration can therefore expose an information problem that remained invisible while the machine simply continued to run.

The operator has authority to act.

It may never have held, and may no longer be able to obtain, the complete design basis needed to explain why the original implementation was acceptable.

## “Foreseen or planned” deserves attention

Regulation (EU) 2023/1230 applies from 20 January 2027 and contains an explicit definition of **substantial modification**.

The current consolidated text defines it as a physical or digital modification made after machinery or a related product has been placed on the market or put into service, which is **not foreseen or planned by the manufacturer**, creates a new hazard or increases an existing risk, and also requires one of two specific forms of additional protective action.

Those conditions concern either guards or protective devices whose integration necessitates modification of the existing safety control system, or additional measures necessary to ensure stability or mechanical strength. \[5\]

The conditions are cumulative.

A controller migration does not become a substantial modification merely because software or hardware changed. A like-for-like migration that preserves the required safety functions at the appropriate PL or SIL will not normally meet the definition simply because the platform is different.

One part of the test nevertheless deserves attention in long-lived automation:

**Was the later change foreseen or planned by the manufacturer of the machinery being modified?**

A controller manufacturer may publish a clear successor path for its own product. That does not necessarily establish what the machine builder foresaw for the complete machine.

Machine-builder instructions can therefore become important lifecycle evidence.

Did they identify permissible controller-replacement routes? Were compatibility limits stated? Was migration anticipated as normal maintenance, or was there a point beyond which the manufacturer expected the safety functions to be reassessed?

There is not yet a mature body of case law, or clear Commission guidance, resolving this particular multi-vendor scenario.

The point should therefore be treated as an emerging legal-engineering question rather than a settled interpretation.

## Future machinery will remember more about its software

The Machinery Regulation also changes the evidence baseline for products entering the new regime.

Annex III requires machinery or related products to collect evidence of legitimate or illegitimate intervention in software or modification of installed software or configuration.

Its control-system provisions require, where appropriate to the circumstances and risks, resistance to reasonably foreseeable malicious attempts that could lead to hazardous situations. The Regulation also requires tracing information concerning interventions and versions of safety software uploaded after placing on the market or putting into service, with a five-year retention period for demonstrating conformity following a reasoned request from a competent national authority. \[5\]

The practical direction is clear:

**New machinery must increasingly be capable of evidencing its own software state.**

That will not prevent obsolescence. It should make a future migration easier to reconstruct because an organisation has a better chance of establishing what software was present, what changed and which safety-related versions formed part of the machine’s operating history.

Many legacy installations begin from a very different position.

## Security-support horizon becomes a buying decision

The Cyber Resilience Act gives this discussion another concrete anchor.

From **11 December 2027**, the CRA applies generally. Manufacturers of products with digital elements must determine a support period taking account of factors including the expected time the product is likely to remain in use. Article 13 also requires the end date of that support period to be clearly specified to the purchaser. \[6\]

For machine builders, that turns security-support horizon into a procurement datum.

A safety-related control platform can no longer be assessed only through current functional capability, price, interfaces and certification. Its specified support period can also be compared with the expected operating life of the machine.

A machine expected to operate for twenty years with a critical digital component whose security-support horizon is materially shorter has a lifecycle decision embedded in the architecture from day one.

That does not make the component unsuitable.

It means the future migration should not come as a surprise.

## A more useful installed-base view

For existing automation estates, I would not begin by cataloguing every PLC in the factory.

Begin with safety relevance, then add security-support status.

The resulting population can be controlled in five states.

**1\. Support status or safety relevance unknown**

Resolve this first. An organisation cannot manage a lifecycle condition it has not characterised.

**2\. Supported and patchable**

Normal vulnerability assessment, qualification and modification procedures remain available.

**3\. Supported, with migration approaching**

Use the remaining support period to preserve expertise, plan shutdowns and establish the successor route.

**4\. Unsupported but mitigated**

Continued operation needs a named risk owner, documented residual-risk acceptance, periodic verification of compensating controls, defined triggers for reassessment and an expiry or replacement date.

A compensating control should itself be checked for unintended effects on the control environment. A firewall rule, network change or communications restriction introduced for cyber reasons can still affect system behaviour and may need to pass through the relevant modification procedure.

**5\. Unsupported with safety-related migration required**

The organisation is no longer managing a patching problem. It is managing a modification to part of the safety-related control system.

There is one additional discipline worth applying to held spares: confirm their firmware state.

Replacing a failed controller with hardware from the storeroom can preserve availability while quietly changing the software configuration that was previously verified.

## Designing for the defensibility date

This changes the design question.

The conventional question is:

**How long will the controller be available?**

The better lifecycle question is:

**How long can we demonstrate that continued use remains acceptable, and what evidence will we need when that changes?**

For new machinery, that means treating security-support horizon, successor-platform strategy, permitted replacement paths and software-state evidence as part of controller selection.

For an installed machine, it means understanding whether an unsupported platform is merely old or whether its remaining operating justification now depends on compensating controls whose effectiveness, ownership and expiry have never been explicitly governed.

The distinction matters because hardware can remain healthy long after the basis for supporting it has changed.

Industrial automation frequently outlives the technology inside it.

The real leadership problem begins when the organisation still owns the machine, still needs the machine and still has authority to decide what happens next, but the evidence needed to defend that decision sits partly in a product history it never fully possessed.

The end of cyber support is therefore not necessarily the end of the machine.

It is the point at which the organisation may have to prove why it believes the machine can safely continue beyond it.

---

# Sources

### Primary regulatory and case sources

**1\. CISA — Schneider Electric Modicon Controllers, ICSA-25-114-01**  
Supports the affected Modicon families, Quantum Safety scope, vulnerability details, compensating measures, end-of-life status, migration recommendation and no-known-public-exploitation statement.  
[CISA — Schneider Electric Modicon Controllers](https://www.cisa.gov/news-events/ics-advisories/icsa-25-114-01?utm%5Fsource=chatgpt.com)

**5\. Regulation (EU) 2023/1230 on machinery**  
Current consolidated EU text supporting substantial modification, Article 18, manufacturer obligations, technical documentation and Annex III software/control-system requirements.  
[EUR-Lex — Regulation (EU) 2023/1230](https://eur-lex.europa.eu/eli/reg/2023/1230/en?utm%5Fsource=chatgpt.com)

**6\. Regulation (EU) 2024/2847 — Cyber Resilience Act**  
Supports the CRA support-period provisions and requirement to specify the support-period end date, with general application from 11 December 2027.  
[EUR-Lex — Cyber Resilience Act](https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R2847&utm%5Fsource=chatgpt.com)

### Standards and professional-practice context

**2\. IEC 62402:2019 — Obsolescence management**  
Lifecycle obsolescence-management framework.  
[IEC — IEC 62402:2019](https://webstore.iec.ch/en/publication/59531?utm%5Fsource=chatgpt.com)

**3\. IEC TS 63074:2023 — Security aspects related to functional safety of safety-related control systems**  
Addresses cybersecurity threats and vulnerabilities capable of affecting safety-related control systems.  
[IEC — IEC TS 63074:2023](https://webstore.iec.ch/en/publication/69228?utm%5Fsource=chatgpt.com)

**4\. ISO/TR 22100-4:2018 — Machinery safety and IT-security aspects**  
Provides machinery manufacturers with guidance concerning IT-security threats that can influence machinery safety.  
[ISO — ISO/TR 22100-4:2018](https://www.iso.org/standard/73335.html?utm%5Fsource=chatgpt.com)