Systemic risks of interconnected infrastructure

+ A ‘systems thinking’ approach is required to increase infrastructure’s resilience to failures and external shocks.

Today’s infrastructure networks are becoming increasingly interconnected. I believe that a ‘systems thinking’ approach is required to increase infrastructure’s resilience to failures and external shocks.

From transport to energy, from water to communications, infrastructure networks are vital to the operation of our day to day lives. When these networks fail, whether in a conflict, natural disaster or due to some other impact, the economic and social impacts can be major. In the worst cases, cascading failures can result in complete social collapse. Therefore, understanding infrastructure interconnections is an increasingly important humanitarian issue.

Historically, infrastructure assets and networks have been considered in isolation in terms of risk. However, networks are becoming more and more reliant upon each other, with increasing interdependencies, which necessitate their continued function. Moreover, the close proximity of infrastructure networks, particularly in cities, can lead to the potential for even a minor failure to result in significant impact to neighbouring assets and networks. A recent example of this was a failed sewer in London which led to the temporary closure of the railway and subsequent delay to many passengers. Another example, which had a truly global impact, occurred in August 2016 when a costly power failure hit Delta Airlines in Atlanta. This resulted in the cancellation of 740 flights over six hours, with every single customer-facing and check-in system unavailable.

As it stands, network intelligence tends to be siloed. Infrastructure owners and operators understandably know their own networks, but often lack insight into external ones they might affect. Yet these external networks might be directly connected and relied upon for their own service delivery. These gaps in operational knowledge and responsibility can lead to systemic risks.

Encouragingly, infrastructure operators and owners, governments and their regulators are becoming increasingly aware of the potential opportunities and threats posed by infrastructure interdependencies. And the research and policy response to infrastructure network issues is growing. The UK Climate Change Risk Assessment 2016 identified cascading infrastructure failures as a key risk and priority, as did the US Department of Homeland Security’s own work. However, every major infrastructure project poses new challenges, so the job of understanding how to assess such interdependencies remains difficult. 

Moreover, embedding comprehensive systems thinking into everyday infrastructure asset management, while addressing the regulatory constraints is challenging. Significant unassessed risks can exist within infrastructure networks. As more and more infrastructure is placed in close proximity, due to population and development pressures, the risks of cascading failure increases. In addition, a range of external environmental threats including climate change, earthquakes and flooding, possess the power to trigger multi-network infrastructure failure.

Software tools and data are building knowledge of infrastructure assets in ways that enable systems thinking. The continuing development and accessibility of open source geographical information systems (GIS) and Building Information Modelling (BIM) means infrastructure networks can be mapped spatially alongside other environmental datasets (e.g. flood risk maps). However, there are issues associated with sharing infrastructure network information, in terms of security, which could potentially reveal certain vulnerabilities or hotspots in a specific country’s infrastructure.

Ultimately, wider adoption of a systems thinking approach could drive improved understanding of infrastructure interdependencies and lead to better informed asset management practices. This in turn will allow infrastructure owners to calculate and plan for the additional external risks to their networks posed by increasing interconnectivity.