Subsea network resilience must be measured by corridor-level risk rather than cable count alone, argues Steve Roberts of EXA. He warns that shared dependencies, geopolitical instability, and repair constraints can undermine perceived route diversity.
Route diversity has long been the benchmark for cable resilience, typically defined by cable count – the idea that more cables mean better protection. This is no doubt a strong strategy: if one cable is damaged, traffic can easily be redirected via an alternative route, which in turn mitigates costly downtime and provides reassurance for both operators and customers.
Yet, in today’s environment, this understanding of route diversity doesn’t go far enough. In fact, multiple cables operating in the same regional corridor have become a new risk in their own right.
Now is the time to rethink resilience and move from counting cables to considering broader corridor risk. True network resilience means assessing corridor-level exposure, repair access, and whether alternative paths can realistically absorb failover traffic.
The illusion of independence
Let’s break this down and unpack the issue with the current definition of network resilience.
The challenge is that the traditional model for resilience assumes that routes are genuinely independent of each other – but that’s not the case. Many cable systems share the same underlying risks, including maritime chokepoints, landing station clustering, overlapping terrestrial backhaul, and common repair dependencies.
Failover between systems that share the same exposure is not true resilience. In a worst-case scenario, this redundancy without independence creates a false sense of security. These shared dependencies become especially problematic when disruption affects entire regions, rather than individual cable systems.
When infrastructure meets instability
With this in mind, today’s geopolitical landscape makes the need for true resilience more important than ever.
Geopolitical instability is increasingly affecting subsea routes, whilst operational constraints are becoming more acute, such as restricted access to open waters, permit delays, and safety risks for repair vessels. Recent disruption in the Gulf illustrates this clearly: the Southern Red Sea became a “no-go zone” for cable repair ships. As a result, any cables damaged in the region, accidental or otherwise, risk remaining down for the foreseeable future.
In this scenario, the risk isn’t just the failure itself, but how long the outage could last due to barriers to repairing the cable. Of course, there are direct implications for network operators, like broken service-level agreements and customer dissatisfaction, but there are also knock-on effects globally: the International Cable Protection Committee estimates that interruptions to subsea systems can have an indirect financial impact of more than $1.5 million per hour.
What real resilience looks like
With this in mind, what practical steps can operators take to build a more realistic resilience model?
The key shift is to begin assessing corridor-level exposure. In other words, the broader geographic and operational environment that cable routes depend on, rather than treating each cable as an isolated asset. There are several factors that a stronger resilience model should evaluate.
Firstly, reassess the concentration of routes and cable landing stations, and where systems physically converge. If multiple systems land in the same regions and rely on the same terrestrial backhaul corridors, there will inevitably be hidden concentrations of risk. A single geopolitical incident, power outage, or environmental event in one location could have a knock-on effect on several systems simultaneously.
In fact, network operators should consider such incidents as part of their infrastructure strategy from the outset. This includes assessing whether certain regions are particularly vulnerable to conflict, sanctions, piracy, or other territorial disputes. Operators need to be able to access cables quickly and consistently to be able to safely carry out repairs if needed. Yet, in some regions, repair timelines can be significantly impacted by political and security constraints, rather than operational or technical factors alone.
Operators should also consider how much repair capacity exists within a region, and how quickly repair ships can be mobilised if needed. Subsea resilience ultimately depends on the ability to restore services quickly, yet only just over 60 repair ships are operating worldwide. These vessels are in high demand, and it’s not uncommon for an operator to face delays while waiting for an available repair vessel. It is therefore important to consider what contingency plans are in place, should conditions deteriorate.
Perhaps most importantly, operators need to test whether backup capacity can genuinely absorb disruption scenarios at scale. Whilst traffic can be rerouted in theory, alternative paths might already be heavily utilised or commercially constrained. Operators therefore don’t need just a physical path, but sufficient operational headroom, too, to maintain service continuity during sustained disruption.
This shift in thinking extends beyond network operators to hyperscalers and high-capacity buyers, too. For them, resilience needs to be reframed as a procurement and architecture decision – not just a network design issue.
Ultimately, in an increasingly complex and tested environment, true network resilience will be defined by how independently cables operate under pressure. Current models are already being stress-tested in real time, and operators that move beyond simple redundancy metrics and adopt a corridor-aware view of risk will be better equipped to anticipate failure, absorb disruption, and maintain continuity when it matters most.
