ISJ hears exclusively from Michael Kolatchev, Principal and Lina Kolesnikova, Senior Consultant of Rossnova Solutions (Belgium) about infrastructure resilience.
Organisations invest extensively in the systems, plans and procedures designed to protect operations and support recovery.
Cybersecurity, backup infrastructure, business continuity arrangements and crisis procedures are all essential components of resilience.
However, major disruptions show that organisations with seemingly comparable levels of preparedness can experience very different recovery trajectories.
Some restore critical operations relatively quickly, while others struggle for considerably longer.
This raises an important question: what determines the difference?
Part of the answer lies in what may be described as the hidden infrastructure of resilience — the relationships, trust and organisational capabilities that enable collective action when established systems and procedures are disrupted.
Resilience requires both stability and flexibility.
Procedures, standards and preventive measures provide structure, but disruption rarely unfolds exactly as anticipated.
Organisations must therefore also be able to adapt when established arrangements no longer fit the situation—a capability that depends not only on formal structures, but also on relationships, trust and patterns of cooperation developed before a crisis occurs.
When formal systems fail
The 2017 NotPetya cyber-attack severely disrupted Maersk’s global operations.
During the recovery, the company had to reinstall approximately 4,000 servers, 45,000 PCs and 2,500 applications.
A surviving domain controller in Ghana provided an unexpected technical foothold from which critical systems could begin to be rebuilt.
The surviving server explains only part of the recovery.
With normal digital systems unavailable, employees had to maintain operations while a global IT infrastructure was being reconstructed.
This required coordination across business units and countries, improvised workarounds and the ability to adapt procedures to circumstances that existing plans could not fully anticipate.
The case illustrates an important distinction.
Technical recovery depended on restoring systems, but operational continuity depended on the organisation’s capacity to act while those systems were unavailable.
Formal structures, responsibilities and procedures remained important, but they had to be complemented by communication, mutual adjustment and local initiative.
Resilience therefore depends not on choosing between structure and adaptation, but on the capacity to move between them as circumstances change.
From capabilities to dependencies
The Maersk experience makes visible something that can be difficult to observe during normal operations.
Coordination, adaptation and continuity depend on patterns of cooperation established before disruption occurs.
Recovery may depend on trust between teams, access to decision-makers or established relationships with suppliers, regulators and other external partners.
The practical question is therefore not simply whether strong relationships exist, but which of them become critical when disruption occurs.
At that point, relationships cease to be merely organisational assets: they become operational dependencies whose weakness or absence can impair response and recovery.
Critical relational dependencies
Organisations routinely identify dependencies on technology, infrastructure, data, suppliers and external services. They are less likely to examine the relational dependencies on which response and recovery may also depend.
A contract may identify a critical supplier, but it does not guarantee effective cooperation during a crisis.
Response may still depend on whether the right people can reach each other quickly, share information and authorise exceptions when normal arrangements no longer apply.
Similar dependencies exist internally. Response may rely on cooperation between IT and operations, security and communications or operational teams and senior management.
Identifying the dependency is therefore only the first step.
Organisations also need to understand what makes that dependency work when normal conditions no longer apply.
Not every people-related dependency is relational. An organisation may depend on an individual because that person maintains a critical relationship; because specialist knowledge is concentrated in that person; or because that person holds the authority required for critical decisions.
These dependencies may overlap, but they are not identical and require different forms of redundancy.
This distinction exposes a particular vulnerability: the relational single point of failure.
Organisations routinely seek to eliminate technical single points of failure, yet access to a critical supplier, regulator or internal capability may depend entirely on one individual’s personal network.
The relationship may be strong while the organisation’s access to it remains fragile.
The objective is not to eliminate personal relationships, which can be extremely valuable during disruption, but to ensure that critical relationships do not remain exclusively personal.
Resilience as a team capability
Relationships can create dependencies, but they can also create redundancy and adaptive capacity.
Crisis performance is rarely the sum of individual capabilities.
A group of highly capable specialists may perform poorly if information is withheld, responsibilities are contested or members have little experience of working together.
Conversely, teams whose members understand each other’s strengths and limitations may compensate for individual weaknesses and adapt more effectively when established procedures fail.
Mick Herron’s ‘Slow Horses’ offers an exaggerated fictional illustration: the deeply imperfect members of Slough House repeatedly function under pressure because their different skills, experience and ability to improvise compensate for individual weaknesses.
Resilience may similarly reside in the interaction between people rather than in the apparent strength of each individual member.
The counterpart to a single point of failure is redundancy.
The principle is familiar from technical resilience: critical functions should not depend on a single pathway.
The same logic can be applied to relationships, knowledge and decision authority.
Critical connections can be maintained by more than one person, knowledge shared and authority delegated.
This does not require everyone to be interchangeable; it requires sufficient overlap and alternative pathways so that the loss of one individual does not become the loss of an organisational capability.
Communication as an adaptive capability
How relationships function under pressure also depends on communication.
As technologies, working patterns and workforce demographics evolve, asynchronous and text-based communication plays an important role in everyday work, particularly among younger employees.
A 2025 UK survey of 1,538 people aged 16–29 found that, among respondents still in education, 30% reported phone anxiety and 38% anxiety about face-to-face small talk.
These preferences should not be interpreted simply as weaker communication skills.
Asynchronous communication has its own advantages: information can be checked, responses considered and exchanges documented.
The resilience question is different: can people change communication modes when circumstances require it?
Crises may require immediate escalation, direct contact with unfamiliar people or decisions based on incomplete information, while the preferred communication channel may itself be unavailable.
An organisation can therefore have robust communication systems while still lacking the ability to adapt how people communicate.
Succession therefore involves more than transferring specialist knowledge.
When experienced employees leave, organisations may also lose informal networks, trusted routes to decision-makers, practised ways of communicating under uncertainty and tacit knowledge of why particular arrangements work as they do.
At the same time, younger employees bring different communication strengths that organisations need to incorporate into crisis arrangements.
The objective is not to preserve one generation’s communication habits, but to maintain a sufficiently broad range of communication capabilities for the organisation to operate effectively when circumstances change.
When the hidden infrastructure is weak
Weak hidden infrastructure can create apparent robustness.
During normal operations, experienced people and established routines may compensate for structural weaknesses, making processes appear more resilient than they are.
Warning signs include unclear decision authority, poor communication across organisational silos, reluctance to escalate bad news and limited coordination with critical external partners.
Crisis arrangements may also depend disproportionately on the availability of a small number of experienced individuals.
Processes may seem mature because particular people know whom to contact, how to resolve bottlenecks and where formal or informal authority sits.
Their absence may reveal that what appeared to be organisational capability was, in fact, personal capability.
These are resilience vulnerabilities, not simply communication or personnel problems.
Testing relationships, not only plans
Crisis exercises often focus on whether plans, procedures and technologies work as intended.
They should also test whether the relationships and organisational connections on which response depends continue to function under pressure.
This broader approach is reflected in large-scale exercises such as ENISA’s Cyber Europe 2026, which brought together more than 5,000 participants in a simulated crisis affecting European rail and maritime systems and tested information sharing, situational awareness and coordination across technical, operational and political levels.
Within individual organisations, exercises can deliberately remove familiar points of coordination: a key decision-maker may be unavailable, a communication channel disrupted or teams required to engage with an external partner through someone other than their usual contact.
Participants can also be required to shift between written, voice and direct communication.
Such scenarios test whether critical knowledge remains accessible and alternative pathways for communication, coordination and decision-making remain available when familiar arrangements disappear.
Friction is valuable evidence.
Delayed decisions, conflicting priorities, blocked information flows or difficulties engaging external partners may reveal weaknesses that a technical test would miss.
An exercise that exposes previously invisible dependencies may ultimately provide greater resilience value than one in which everything proceeds according to plan.
Resilience beyond organisational boundaries
The hidden infrastructure does not stop at the organisational boundary.
Supply chains, outsourced services, shared platforms and technology ecosystems mean that response and recovery increasingly depend on external organisations and capabilities beyond direct managerial control.
The 2025 Collective Cyber Resilience Index, based on a survey of 200 US federal and private-sector cybersecurity decision-makers, reflects this shift: 85% considered cross-organisational collaboration essential to their defence strategies, while 72% believed that inconsistent security practices across partners may pose a greater threat than cyber-adversaries themselves.
Identifying a critical supplier or service is therefore not enough.
Organisations need to understand how information, coordination and decision-making will work across organisational boundaries when those dependencies are under pressure.
These arrangements also require maintenance.
Suppliers change, organisations restructure, responsibilities move and experienced people leave.
Communication practices evolve as well.
A relationship or coordination mechanism that functioned during the last crisis may no longer be available during the next one.
Like any infrastructure, the hidden infrastructure of resilience can deteriorate if it is not maintained.
Conclusion
Organisations routinely identify critical technologies, suppliers, systems and services.
They should apply similar scrutiny to the less visible dependencies that determine whether those resources can actually be mobilised under pressure.
This means identifying critical relational dependencies, recognising relational single points of failure and testing whether the organisation can still access the knowledge, relationships and authority it needs when familiar arrangements fail.
A useful question is therefore not simply whether an organisation has strong relationships, but where recovery would fail if trusted relationships were absent.
Technology, plans and procedures remain indispensable. But resilience also depends on an infrastructure that is harder to see: the relationships and organisational capabilities that enable collective action when normal arrangements fail.
The hidden infrastructure of resilience must therefore be identified, developed, tested and maintained before it is needed.