- INTRODUCTION: Where Great Transformations Converge
Various reports from leading strategic consulting firms anticipate that the coming decades will be marked by an extraordinary expansion in infrastructure investment. Estimates vary depending on which assets and sectors are included, but they converge on investment needs amounting to hundreds of trillions of dollars, driven by urbanisation, the energy transition, digitalisation, climate adaptation, and the maintenance and renewal of existing infrastructure. This has led to growing references to a new infrastructure supercycle, or even an infrastructure hypercycle.
The sheer scale of these figures, however, can lead to an incomplete interpretation of what is happening, because the more relevant question is different: it is not so much the scale of the investment that will be required as the changing nature of the problem itself. Because that change may give rise to new ways of creating value, new actors and, probably, a new dimension of civil engineering.
Several of the great transitions—indeed, I would argue the most important ones—that we are experiencing, and that are redefining the contemporary economy and geopolitical landscape, are converging on the same material foundation. The energy transition requires generation capacity, electricity grids, storage, critical minerals and new logistics chains. The digital revolution and artificial intelligence depend on semiconductors, data centres, electricity, water, rare earths and connectivity. Climate adaptation requires the simultaneous transformation of urban, water, energy and transport systems. And growing geopolitical competition has turned ports, submarine cables, electricity grids, supply chains, computing capacity and semiconductor manufacturing into assets of strategic and critical importance to national security.
Phenomena that we usually analyse as separate transformations thus reveal a shared infrastructural foundation. They all depend on our ability to build, connect and govern increasingly interdependent physical and digital systems. Perhaps this is the most useful way to understand twenty-first-century infrastructure: not as a collection of sectors, but as the system of systems that enables many of the great transformations of our time.
This interdependence is changing the very nature of infrastructure. For much of the twentieth century, the problem could largely be framed in sectoral terms: providing transport, energy, water or telecommunications capacity where it was needed. That logic remains essential, particularly where basic infrastructure deficits persist. But it is no longer sufficient to describe the emerging challenge. A growing share of the economic and social value of infrastructure depends not only on the performance of an asset considered in isolation, but on its ability to function as part of broader systems.
This transformation has profound consequences. It changes the nature of risk, because interdependencies allow disruptions to propagate across systems. It changes the relationship between infrastructure and territory, because access to energy, connectivity, logistics or computing capacity can determine where new economic activities are located. It changes the geopolitical dimension of infrastructure, because control over particular nodes, networks and interfaces creates new forms of dependence and power. It even changes the meaning of time: in a context of technological acceleration and competition between countries, the speed at which a society can deploy and connect infrastructure may determine its ability to capture new economic opportunities.
To speak of a new Age of Infrastructure, therefore, is not simply to claim that we will need to build more. What is emerging is a qualitatively different form of infrastructure: more interdependent, more strategic, and progressively more difficult to understand and govern through sector-based structures. This shift requires us to reconsider how we understand value, risk, sovereignty, territory, equity, time and the corporate capabilities associated with infrastructure.
2. SHIFT: A New Concept of Infrastructure
For much of the industrial era, infrastructure could be conceived and managed primarily as a set of distinct sectoral systems: transport, energy, water or telecommunications, each with its own technologies, institutions, financing models and regulatory frameworks. This representation was never entirely accurate—infrastructure systems have always depended on one another—but it was sufficiently useful to organise their planning, construction and management. Today, it is becoming increasingly inadequate.
Contemporary infrastructure increasingly resembles a common enabling layer on which the economy and society operate. Industrial production, trade, mobility, public services, communications and, increasingly, the digital economy itself depend on specific combinations of energy, water, transport, connectivity and computing capacity. In this sense, it may be useful to think of infrastructure as the operating system of civilisation.
The analogy also allows us to broaden the traditional definition of infrastructure. This operating system is no longer exclusively physical. It consists of material assets—roads, electricity grids, ports, pipelines or data centres—but also of software, sensors, data, protocols, standards, institutions, regulation, organisations and people. The boundaries between physical, digital, human and institutional infrastructure are becoming progressively more permeable.
This shift has a particularly important consequence for engineering. For much of its history, civil engineering has developed an extraordinary capacity for engineering assets: designing bridges, dams, roads, ports, railways or networks to deliver defined levels of performance with high standards of safety and reliability. That capability will remain indispensable. But a growing share of infrastructure performance today resides not exclusively in each individual asset, but in how assets, networks and the institutions that govern them interact.
The challenge therefore shifts, at least in part, from engineering assets to engineering interactions. It is no longer enough to ask whether each component functions correctly. We must ask what happens at their interfaces, what dependencies we are creating, what additional capabilities a connection enables, and how a disruption may propagate through the system.
There is an additional difficulty. We are building assets with useful lives spanning several decades for a world whose technological, climatic, economic and geopolitical evolution is becoming increasingly difficult to anticipate. Infrastructure built today may still be operating in 2070 or 2080, under conditions that we can hardly specify with precision from where we stand today.
This requires us to reconsider a planning culture based primarily on prediction, efficiency and optimisation. Not because these principles are no longer valid, but because excessively optimising a system for the future we consider most likely can reduce its ability to respond to different futures. Twenty-first-century infrastructure will require efficiency, but also redundancy, diversity, flexibility and optionality. Maintaining alternatives that appear unnecessary under normal conditions can prove extraordinarily valuable when our forecasts fail.
The design problem therefore changes in nature. It is no longer simply about finding the optimal solution for a set of expected conditions, but also about preserving the system’s capacity to evolve when those conditions change. This implies moving, to some extent, from engineering a system to steering an evolving system.
3. VALUE: Integration Becomes a New Source of Value
Anyone following the infrastructure market closely may have noticed a series of moves by major companies that do not fit easily within traditional sectoral boundaries. Firms historically associated with construction, engineering or infrastructure operation are expanding into energy, data centres, digital technologies, advanced manufacturing and other adjacent fields. The most immediate explanation is diversification: companies are entering attractive new markets, spreading risk and positioning themselves in sectors expected to grow rapidly over the coming decades. In many cases, that may indeed be what is happening.
But there is another possible interpretation. If infrastructure is shifting from a collection of largely independent assets towards increasingly interdependent systems, some of these moves may reflect not simply a diversification of portfolios, but a growing logic of integration. This may be creating space for a new kind of actor: the Infrastructure Integrator, capable of orchestrating multiple infrastructure systems, technologies, institutions and sources of capital to jointly produce a capability that none could provide independently. It is therefore not simply a large contractor or a diversified company operating across multiple businesses. The distinction lies not in how many sectors it occupies, but in the logic that connects them.
The difference becomes clearer with a limiting example. Possessing separate expertise in energy, data centres, semiconductors, engineering and construction constitutes a diversified portfolio. Using that knowledge to anticipate how advances in chips will alter power density and cooling requirements, how those changes will shape data-centre design, what generation and grid capacity will consequently be required, and how all these elements can be synchronised to deliver computing capacity sooner constitutes integration. The competitive advantage no longer resides only within each capability, but increasingly in the knowledge of the interfaces between them and in the ability to orchestrate them as a system.
This does not mean that integrators will replace the infrastructure ecosystem that already exists. Specialist manufacturers, technology companies, engineering firms, contractors, operators, investors and public institutions will remain indispensable. Rather, increasing interdependence may create the need for an additional layer of actors operating across traditional sectoral boundaries, connecting and coordinating the capabilities of the existing ecosystem.
Behind the emergence of these actors lies a corresponding economic logic: when coordinating complementary infrastructure systems makes it possible to create greater capability, create it sooner, or deliver it at a lower systemic cost than by optimising those same systems independently.
4. GOVERNANCE: Complexity Must Be Governed Differently
Infrastructure systems are becoming integrated faster than the institutions responsible for planning, financing, regulating and governing them. Ministries, regulators, companies, budgets and appraisal procedures administratively divide a reality that increasingly operates as a system.This gives rise to a complexity mismatch: a gap between the complexity of the system we are trying to govern and the capacity of our institutions to understand and manage its interdependencies.
There is also a deeper difficulty: we have not only organised our institutions along sectoral lines; we also tend to think in sectors. Much of engineering, economics and public administration has developed by breaking complex problems down into manageable parts, establishing relatively clear relationships between cause and effect, and optimising the variables that can be observed and controlled. This approach has been extraordinarily powerful. But it has limitations when applied to systems characterised by feedback loops, delays, dependencies, emergent behaviour and unintended consequences.
Governing this complexity does not necessarily mean creating larger institutions or concentrating all decisions in a single centre. The challenge is rather to preserve specialisation where it generates knowledge and technical capability, while building mechanisms for integration where decisions interact.
This brings infrastructure governance closer to what is described as polycentric systems: structures in which multiple centres of decision-making retain autonomy while developing rules and mechanisms for coordination. Applied to twenty-first-century infrastructure, this means building an architecture capable of combining distributed knowledge with integrative capacity. We do not need to eliminate institutional boundaries; we need to make them more permeable where the interdependencies of the system require it.
Standards will be one of the invisible infrastructures of this integration. The interoperability of physical and digital systems, data exchange, communication protocols, cybersecurity, payment and access systems, and compatibility between technologies will require shared frameworks that allow operators and services to connect without necessarily imposing a single technological or institutional model.
But standards are not neutral. Those who define the interfaces can shape the evolution of the systems built upon them. An apparently technical decision about a communication protocol, data architecture, charging system or digital interface can create network effects, switching costs and technological trajectories that are difficult to reverse. The contest over the infrastructure of the future will therefore be fought not only over who finances, builds or controls the assets, but also over who defines the rules by which those assets connect. Interfaces are points of power.
The same logic requires us to broaden the perimeter of governance. An infrastructure system may simultaneously depend on municipalities, regions, national governments, regulators, public companies and private operators. But they are now being joined by actors that until recently were barely part of the infrastructure system: digital platforms, hyperscalers, autonomous vehicle operators, mobility companies, energy aggregators, cloud providers, telecommunications companies and data managers. Power over infrastructure is therefore becoming distributed among actors that do not necessarily share the same objectives, time horizons, incentives or accountability mechanisms.
The role of the citizen is changing as well. Digitalisation, distributed generation, batteries, electric vehicles, shared mobility and, increasingly, autonomous systems are blurring the boundary between those who use infrastructure and those who participate in its operation. Millions of individual decisions can collectively determine peaks in electricity demand, urban congestion, storage requirements, network utilisation or the availability of capacity. The user is no longer simply the final recipient of a service but, in certain systems, an active variable in how that system operates.
In a world of increasingly interdependent systems, capability will no longer depend only on how much capital can be mobilised or which technologies can be acquired. It will also depend on how much complexity we are able to understand, integrate and govern.
5. TEMPO: Timing Becomes Strategic
Time is no longer simply a management variable; it is becoming a strategic one. The speed of technological, climatic and geopolitical change increasingly contrasts with the long cycles required to plan, approve, finance, manufacture and build infrastructure. Delay creates two distinct risks: exposure to threats and exclusion from opportunities. Arriving late to climate adaptation increases our exposure to events for which we are unprepared; arriving late to electricity grids, energy generation, connectivity, water, logistics capacity or land availability may mean losing investment and the economic ecosystems that develop around it.
This temporal dimension becomes even more important in the new geopolitical context. Technology is once again a question of power, and so is the infrastructure that enables its development. Artificial intelligence illustrates this transformation with particular clarity. Its development may appear essentially digital, but it rests on a profoundly physical reality: data centres, abundant and reliable electricity, power grids, water, fibre-optic networks, land, cooling systems, logistics chains and the industrial capacity to manufacture advanced equipment. The digital frontier has a physical foundation. The speed at which a country can deploy this material foundation will increasingly determine the speed at which it can expand its computing capacity and develop its digital economy.
But the competition will not simply be about having the infrastructure required to produce artificial intelligence. A second race may prove even more important: integrating AI into the functioning of the economy and infrastructure itself. Electricity grids capable of anticipating demand and managing millions of distributed resources; transport systems coordinating vehicles, roads and public transport in real time; water networks capable of detecting leaks and forecasting demand; ports and logistics corridors coordinated through data; infrastructure able to anticipate its own maintenance needs; and cities capable of dynamically managing energy, mobility and public services. Infrastructure will progressively cease to be merely the physical foundation of the digital economy and become one of its principal fields of application.
Advantage will not necessarily belong to whoever invents a technology first, but also to whoever can deploy it sooner, integrate it at scale and learn more rapidly from its use. Infrastructure thus becomes a mechanism for technological diffusion and cumulative learning. In certain sectors, arriving late does not simply mean starting later; it means starting from a structurally weaker position. The ability to narrow that gap—to decide, approve, finance, build, integrate and learn—may become one of the most important strategic capabilities of states and economies.
Prudence, rigorous appraisal and institutional stability remain essential when making decisions that commit enormous amounts of capital and shape territories for decades. But there is also a cost to waiting. The temporal dimension of infrastructure therefore extends far beyond the speed of execution. It is about acting fast enough not to be left out of the future, while retaining enough flexibility not to become trapped by our own vision of that future. Advantage will not lie in accurately predicting the decades ahead. It will increasingly lie in building systems capable of thriving even when our predictions are wrong.
6. PURPOSE: Expanding Human and Societal Agency
And what is all this ultimately for? We should never forget that the ultimate purpose of infrastructure is not to connect systems, but to protect people and to expand human agency.
This is not a new mission. Large shared systems for water, sanitation, energy, transport and communications free people from having to meet essential needs individually and multiply what they can do with their time, knowledge and capabilities. In this sense, infrastructure has been an extraordinary collective technology for expanding human agency.
The new generation of infrastructure could take this logic much further. But precisely for this reason, we need to qualify the apparently straightforward relationship between infrastructure and agency. Until now, what we delegated to infrastructure was primarily physical effort and time. Increasingly, we will also delegate observation, interpretation, prediction and decision-making. A smart electricity grid, an autonomous mobility system, an algorithmically managed city or public services mediated by artificial intelligence do not simply provide capabilities that we then use freely. They observe, anticipate, recommend, prioritise, optimise and, in some cases, decide. We are crossing an important frontier: from infrastructures that enable human action to infrastructures that increasingly participate in human agency.
This introduces a human dimension to resilience. For decades, we have designed physical redundancy: additional capacity, alternative routes, storage and backup systems. In increasingly autonomous infrastructure, we will also need a degree of cognitive redundancy: people and organisations that retain sufficient knowledge, experience, judgement and capacity for intervention to understand the system, challenge it and regain control when necessary. Maintaining these capabilities may appear inefficient while automation works perfectly, just as an alternative route or reserve capacity appears inefficient when it is never needed. But that is precisely where its value lies. Preserving human agency may become a resilience strategy.
There is, however, a second and even more important social dimension. Access to competitive energy, computing, data, artificial intelligence, knowledge and automation may determine which companies can compete, which researchers can make discoveries, which workers can increase their productivity and which territories can participate fully in the new economy.
This requires us to reconsider what equity means in infrastructure. For much of the twentieth century, one of the great objectives of infrastructure policy was to universalise access to essential services. In the twenty-first century, we will need to ask whether some of these new enabling resources—advanced connectivity, computing, artificial intelligence and certain forms of knowledge infrastructure—require a similar ambition. The challenge will no longer be simply to ensure that people are connected, but to prevent a situation in which one part of society has permanent access to systems that multiply its capabilities while another is structurally excluded from them. Infrastructure can accelerate growth and, at the same time, accelerate divergence.
Social cohesion is not merely a desirable consequence of development. In societies undergoing rapid transformation, it can become a strategic asset: it shapes their ability to accept change, sustain long-term investment, share risks and maintain trust in the institutions managing that transformation.
There is therefore also a social contract surrounding the new infrastructure. The decisions we make will determine who pays for the transition, who receives its benefits, which territories concentrate opportunities, which capabilities we transfer to machines and which we choose to preserve in people. Not all these decisions can be resolved through technical optimisation. They require legitimacy, participation and trust. Integrating infrastructure without integrating the societies that depend on it would be an incomplete—and potentially unsustainable—form of integration.
The point, therefore, is not to add a social dimension to what is essentially a technological agenda. It is to define the mission of that agenda with its social purpose firmly in view. Over the coming decades, we will invest extraordinary amounts in transforming the material and digital foundations on which our civilisation operates. Those decisions will shape the competitiveness of economies and the power of states, but also the autonomy of citizens, the distribution of opportunity and the cohesion of our societies.
7. ACTION: World Infrastructure Forum
We are entering a new age of infrastructure, one in which the systems that underpin our economies and societies will play an increasingly vital role in shaping prosperity, resilience, sovereignty and human progress. Closing the infrastructure gap will require enormous amounts of capital, technology and engineering capability. But it will also require something increasingly scarce and valuable: the collective capacity to understand interdependence, govern complexity, act in time and turn integration into broadly shared progress.
This is the intellectual rationale for the World Infrastructure Forum. Its ambition is not simply to bring together the largest organisations or to ensure that every sector is represented. It is to create a place where the boundaries that still structure the infrastructure ecosystem can become more permeable—and, in doing so, increase its collective capacity to integrate. A neutral space where interdependencies can be identified before they become bottlenecks, where decisions operating on different timescales can be better synchronised, where knowledge can move across sectors, and where we can ask not only what we can build, but what capabilities we want to créate and for whom. The ambition of the World Infrastructure Forum is therefore to increase the integration capability of the global infrastructure ecosystem, so that better infrastructure ultimately expands the possibilities of the people and societies it is meant to serve.
José Miguel Atienza
Director
Escuela de Ingenieros de Caminos, Canales y Puertos
Universidad Politécnica de Madrid (UPM)
https://worldinfrastructureforum.org