Legacy strategy

One Router Grounded America. How Do You Modernise Without Bringing Everything Down?

America’s aviation outage illustrates a challenge facing every CIO: how to reduce dependence on ageing technology while protecting the critical services that depend on it.

4 min read

Mark Lister

A router being installed at a remote location seems an unlikely trigger for nationwide aviation disruption. Yet, according to the US Department of Transport, work to replace ageing communications infrastructure caused a router to flood the air traffic control network with connection attempts. The precise cause remained under investigation but the consequences extended far beyond the building where the work took place.

The Financial Times’ account, republished by To Vima, presents a troubling picture of dependence on old technology and repeated appeals for the money to replace this critical infrastructure. Financial Times / Tovima; but the real scope of “updating the US aviation technology’s legacy” can’t be judged by the size or cost of one component being replaced.

That concern echoes the central theme from Mitra’s recent UK CDIO Advisory Board: organisations often commit to transformation before they sufficiently understand the systems, processes and dependencies involved. Important discoveries then arrive after the budget, timetable and business expectations have been set and, by that point, learning becomes expensive.

The old estate holds the kernel of the answer

Legacy systems embody years of business decisions. Their code contains pricing rules, contractual commitments, operational exceptions and controls. Their data records how the organisation has actually worked, including the inconsistencies and workarounds that formal process diagrams can miss.

Some of that inheritance remains valuable but some of it will now be redundant, getting in the way and actively constraining the business. It’s a technical debt that needs to be managed “out”. At Mitra we believe that precise understanding of the existing estate is a critical part of the transformation itself. Examining the code and tracing how data moves through and between systems helps reveal what the organisation depends on, where knowledge is missing and where change could have unexpected consequences.

Code alone cannot explain why a rule still matters and data alone cannot establish whether a process should continue. Business owners and experienced operational staff must help interpret the evidence and decide what deserves preservation, simplification or retirement. The US aviation industry needs a proper plan to reduce the likelihood of anything like this happening again.

Make uncertainty visible before making the commitment

Technical debt becomes more manageable when leaders connect it to business risk and exposure. They need to clarify which services would be affected by a failure, how quickly they could recover, where support or skills are disappearing and which dependencies make change difficult.

Age matters, but it is an incomplete guide to priority. A well-understood older application with effective controls may present less immediate exposure than a newer service whose connections and recovery arrangements are poorly understood.

The first step to knowing “as much as possible” should therefore establish an evidence-based view of the estate and its uncertainties to help produce a prioritised decision map: what needs urgent attention, what can safely remain, what can be simplified and what requires further investigation.

AI can accelerate parts of this discovery by helping teams interpret code, identify candidate business rules and assemble documentation for review. Its findings still need validation against actual behaviour, operational evidence and business knowledge. A convincing explanation is a hypothesis until it has been checked by the expert humans who need to make the tough decisions.

Progress through changes you can prove

Once the estate is better understood, the route forward can become better controlled. An organisation can retain a stable component, isolate a fragile dependency, improve access to valuable data, rebuild a constrained function or replace an unsupported platform: different parts of the estate may justify different treatments.

Mitra’s approach to progressive modernisation is to make those choices through manageable increments, with explicit review milestones along the way. Each stage should deliver useful capability or evidence that improves the next investment decision.

Before each release, teams need evidence that essential behaviour has been preserved, data remains consistent and recovery is practical. Depending on the service, that may require representative workload testing, comparison of old and new results, a limited initial rollout or rehearsed recovery procedures.

Rollback also needs scrutiny. Restoring an earlier software version may be straightforward; but reversing transactions or reconciling data written during a failed release may end up being much harder.

These disciplines cannot guarantee failure-free change but they help make uncertainty explicit, limit exposure and give leaders defensible grounds for proceeding.

Retire the dependence as well as the technology

Progressive modernisation needs a destination. Adding new services around old systems can deliver early benefits, but prolonged coexistence can also increase cost, complexity and support obligations.

The pursuit of simplification means every transitional arrangement should have an owner, a purpose and clear conditions for retirement. Otherwise, today’s route out of technical debt becomes tomorrow’s additional layer of new debt.

The knowledge recovered during discovery should also remain as an accessible organisational asset. Business rules, dependency maps, data definitions and the reasoning behind decisions need to be available and current and their value should survive the departure of an individual, a delivery partner or a technology supplier.

For boards, this changes what progress looks like. Applications migrated and code rewritten are useful delivery measures but stronger evidence of success includes fewer critical dependencies, tested recovery, less reliance on scarce individuals, simpler processes and lower effort to make the next change.

The aviation disruption should sharpen the resolve to address ageing infrastructure. It should also sharpen the questions asked before touching it. You can’t remove all the risk of modernisation, but you can learn from what already exists, preserve what the business still needs and progressively retire what holds it back.

That is how understanding old code and data becomes a practical route to a more resilient future.