The quiet infrastructure crisis

Industry, healthcare and defense rely on infrastructure whose lifecycle is often not systematically planned. BRESSNER Technology explains why hardware end of life is a matter for company leadership.

In the conversations we have every day with engineers, IT leaders, procurement teams and managing directors, one issue comes up increasingly often, even if nobody calls it by name: the hardware running critical systems is getting older. Very few companies have a clear answer to what happens when it reaches the end of its availability.

This is not a criticism. It is an observation. For a long time, end of life was an IT matter: a routine task handled somewhere within the department, without involving company leadership. That is changing, faster than many companies realize.

Planning digitalization around hardware that will reach EOL in three years means planning twice.

When EOL stops being routine

The conventional scenario is simple: a manufacturer discontinues a model, procurement orders its successor, and operations continue. In office environments, this usually works. In industrial plants, healthcare facilities and defense systems, it often does not, because the systems cannot simply be replaced.

A production line that has run on a particular hardware platform for eight years has interfaces, certifications and integration points tied to that exact platform. In an MDR-compliant medical device, changing the hardware platform may require reassessment and, where applicable, requalification. Without a documented equivalent replacement, the product’s conformity may be at risk. A defense system with a planned service life of 15 years faces a serious problem if the manufacturer declares its computing component EOL after seven years and the only available alternative is outside Europe.

These are not extreme scenarios. They are inquiries we receive regularly. In almost every case, the problem could have been managed if it had been identified earlier.

A wave we can already see coming

The generation of hardware installed in industrial plants, healthcare facilities and defense systems between 2010 and 2018 is approaching the end of its availability. One example from the embedded sector is the Intel Atom E3800 family. Introduced in 2013, the platform was used in long-life embedded applications and illustrates how a suitable system platform can become a lifecycle concern years later. Before any new long-term project using platforms of this generation, support, availability and spare-parts supply must now be checked in detail. This is not a prediction: these are product lifecycles evident from manufacturers’ information and market observations.

At the same time, many companies are running digitalization projects, AI pilots and Industry 4.0 initiatives on this very infrastructure. According to a representative survey by the digital industry association Bitkom, one in three companies in Germany already uses AI, almost twice as many as a year earlier. Almost one in two plans to use it or is discussing doing so (Bitkom, September 2025; sample: 604 companies with at least 20 employees). New requirements, new software and new processes are being built on hardware designed for other demands, whose availability nobody has systematically planned. This combination will become apparent in many companies over the coming years: through a series of avoidable problems rather than a single major crisis.

The most expensive hardware decision is the one you never make.

Why this belongs on the board’s agenda

Hardware lifecycle management sounds like an IT issue. On closer inspection, it concerns operational continuity, investment security and, depending on the industry, regulatory compliance. NIS2, requirements for critical infrastructure and the MDR increase the pressure to plan responsibilities, documentation, the ability to update systems and critical-system lifecycles on a sound basis. Company leadership therefore needs to know which hardware its critical systems use, how long it will remain available and what alternatives exist when it reaches EOL.

This does not mean that boards suddenly have to manage product lifecycles themselves. It means that the questions of which hardware runs our critical systems, how long it will remain available and what our plan is when it no longer is need answers that extend beyond the IT department.

What companies can do now

The first step is usually the simplest and the one most often skipped: an inventory. Which hardware runs which critical systems, and when will it reach the end of its availability? Most companies already have this information; it is rarely brought together.

The next step is lifecycle planning integrated into investment planning, rather than IT budgeting. For critical components, it is also worth establishing early whether a second source of supply exists, ideally in Europe and with documented qualification. Companies that resolve this in 2026 will have options in 2028. Those that wait until 2027 will face pressure.

The real issue

The quiet infrastructure crisis is more than a technical problem. It stems from a strategically important question being systematically left unasked because it looks like an operational one.

More than three decades of working at the interface between manufacturers, integrators and operators have taught us one thing: the problems we see today almost always gave early warning. They were not recognized as problems because they had not yet become problems. That is the moment to act, before pressure determines which options remain.

Frequently asked questions

Image source: AI-generated (September 2026)

Translated from the German original using artificial intelligence. The German version is authoritative.