OCFor years the infrastructure conversation has centred on one question: where will the investment come from? That is no longer the most important question.
The UK Infrastructure Pipeline now contains 734 planned projects representing £718bn of public and private investment over the next decade. Energy alone accounts for £365bn of it. Delivering the pipeline is expected to require between 629,000 and 706,000 workers every year for the next five years.
The opportunity is enormous. So is the delivery challenge.
And it is worth noting that the Public Accounts Committee has already warned that the authority responsible still lacks full oversight of the government's infrastructure portfolio, having previously described the pre-NISTA pipeline as not credible. That is not a funding problem. That is an execution problem, stated by Parliament, before a spade has gone in the ground.
01Infrastructure is becoming a system, not a collection of projects
Consider what is happening around artificial intelligence. The expansion of AI is creating demand for data centres. Data centres require electricity, grid connections, land, cooling and water. The same electricity networks are being asked to support electrification of transport, heating, industry and new housing. The same land may be required for homes, logistics and energy infrastructure. The same engineering skills are needed across all of it.
A data centre is no longer simply a technology investment. A housing programme is no longer simply a property programme. An energy project is no longer simply an engineering programme. Each depends on the capacity and timing of several other systems.
That is where otherwise credible investment plans will struggle.
02The real constraint is organisational capacity
There is a tendency in major programmes to assume that once funding, technology and contractors are secured, execution follows.
Experience suggests otherwise. I have spent twenty years inside these programmes, at Shell, Chevron, BP and major UK utilities, and large transformations rarely fail because nobody had a strategy. They struggle because hundreds of individual decisions, dependencies and responsibilities have to work together over several years.
The difficult questions are more practical than strategic. Who owns the outcome? Are the operating model and governance clear? Can the organisation decide quickly enough? Are engineering, commercial, technology and operational teams working from the same assumptions? Are dependencies understood before they become critical?
And most importantly: is the organisation actually capable of absorbing the change it has commissioned?
These questions are less exciting than announcing another billion-pound programme. They are also where value is won or lost.
03What this looks like in practice
When I led the process implementation team on Shell's Collaborative Working Environment programme, across their African assets, Kazakhstan and ONEgas, the estimated annual benefit was over US$75m in production efficiency. Well restart times fell from six days to under three. Travel to the field dropped by 15%.
None of that came from new equipment. It came from changing how people worked, who decided what, and when.
More recently, on a remote operations study for an offshore gas facility, we mapped every manual intervention on the platform. Around 80% of maintenance hours were going on corrective work: fixing things after they broke. The prize was not a new system. It was shifting that towards predictive and preventive, and building the governance to make it stick.
And on a major UK water utility's asset management transformation, the work before the new enterprise system could even be designed was aligning operational and maintenance processes across three regions that had been separate companies. Without that, the technology would simply have automated three different ways of doing the same job.
04AI makes the execution problem more urgent, not less
Artificial intelligence will improve parts of infrastructure delivery. Planning, engineering analysis, asset management, predictive maintenance, procurement and scheduling will all benefit.
The danger is believing that technology removes the need to fix the underlying organisation.
Putting sophisticated AI on top of fragmented processes, unclear accountability and poor data simply allows an organisation to make bad decisions faster.
Before asking where AI can be deployed, leaders should ask which decisions, processes and operational outcomes they are trying to improve. Technology should enable the operating model. It should not become the operating model.
05There is also a resource question nobody can ignore
AI may appear digital, but the infrastructure supporting it is decidedly physical. Servers require buildings. Buildings require land. Computing requires electricity. Cooling requires infrastructure and, depending on the technology, significant water. Networks require grid capacity.
Which creates a contradiction. The technology expected to make economies more productive is simultaneously increasing demand on the physical systems already struggling to accommodate growth.
This is why energy strategy, digital strategy, infrastructure strategy and industrial strategy can no longer be developed independently. They are becoming the same conversation.
06What leaders should do differently
The organisations best positioned for the next decade will treat execution capability as an investment in its own right. That starts with six disciplines.
- Understand the whole system before committing to individual solutions. Map the dependencies across infrastructure, regulation, technology, people, suppliers and operations.
- Establish genuine accountability. Programmes become vulnerable when everybody participates and nobody owns the outcome.
- Design for operations from the beginning. The people who will run the asset should influence its design long before commissioning.
- Challenge organisational capacity honestly. An organisation running today's operation at full stretch cannot absorb a major transformation without consequences.
- Use technology selectively. AI and data should solve identifiable operational problems, not become programmes searching for justification.
- Maintain independent challenge. Large programmes need people prepared to ask uncomfortable questions before those questions become expensive.
07From ambition to execution
The UK is not short of infrastructure ambition. The pipeline is substantial, private capital is interested, and new technologies are creating real opportunities.
The harder question is whether organisations can deliver what they are planning at the speed now expected.
That is where the next competitive advantage will emerge. Not from more capital. Not from better technology. Not from another strategy document. It will come from organisations capable of connecting strategy, people, process, technology and delivery into one coherent system.
Because infrastructure does not create value when it is announced. It creates value when it works.