03
^{CAPITAL PROJECTS}
Construction: the critical path towards capital efficiency

^{CAPITAL PROJECTS}

Construction: the critical path towards capital efficiency

How the industry must reinvent itself to deliver the infrastructure the world needs

In the most recent World Energy Investment report, IEA projected global energy spending to reach $3.4 trillion in 2026, a 5% increase over the previous year.¹ Analysis also showed high investment in clean energy, with two dollars flowing to renewables, nuclear, grids, storage, low-emissions fuels, efficiency and electrification for every one dollar invested in oil, natural gas and coal.²

But investment alone will not deliver the energy transition.

$3T

Expected global energy investment in 2026 (IEA)

5%

YOY increase between 2025-2026 (IEA)

¹ https://www.iea.org/reports/world-energy-investment-2025
² https://www.iea.org/reports/world-energy-investment-2025

As demand surges and spending accelerates, the energy industry's greatest challenge has shifted from funding infrastructure to building it. Grid connections, transmission upgrades, interconnectors, and other critical assets must all be delivered at a speed and scale that the current construction delivery model cannot support.

As a result, construction has become one of the defining constraints of the global energy transition. Overcoming it requires a fundamental reinvention of how infrastructure is designed, engineered and delivered.

Here we explore how the construction industry must evolve toward a fundamentally different model—one that is outcome-led, alliance-based, digitally integrated, and focused on delivering value across the entire asset lifecycle—and why doing so will not only support the energy transition, but also help fulfill the sector’s own growth ambitions and meet capital efficiency expectations.

Figure 1: IEA: Global investment in clean energy and fossil fuels, 2015-2025

^{IEA (2025), Global investment in clean energy and fossil fuels, 2015-2025, IEA, Paris3}

³ https://www.iea.org/data-and-statistics/charts/global-investment-in-clean-energy-and-fossil-fuels-2015-2025

Figure 2: IEA: Clean energy investments by region, 2025⁴

China	$627 billion
European Union	$386 billion
United States	$400 billion
Other emerging markets and developing economies	$300 billion
Other advanced economies	$175 billion
Japan and Korea	$103 billion
India	$100 billion

What needs to change in construction. (And why the answer isn’t just AI.)

Despite unprecedented demand for new energy infrastructure, construction productivity has increased by just 0.4% annually over the past two decades, compared with around 3% in manufacturing and 2% across the broader economy.⁵

However, in recent years, momentum is beginning to build. In 2024, more than 64% of construction firms were using or piloting AI to improve efficiency and reduce costs.⁶

While this growing adoption is encouraging, technology alone, especially when deployed as point solutions, won't solve the industry’s overarching productivity challenge. Instead, construction organizations need to rethink how they operate, how people and partners work together, and the foundations that support them. Here we offer five interconnected changes that can help the construction industry scale delivery and support the energy transition.

⁴https://www.iea.org/reports/world-energy-investment-2025/executive-summary
⁵ https://www.vsimple.com/blog/mckinsey-study-shows-the-construction-industry-is-in-a-productivity-rut
⁶ https://www.startus-insights.com/innovators-guide/construction-industry-trends/

Change 1: The operating model

Construction organizations have traditionally been designed to manage individual projects, with governance, decision-making and resources optimized for project delivery rather than portfolio-wide outcomes. However, that model was never intended to support the unprecedented scale of investment now flowing into the industry, particularly across energy infrastructure.

Rather than managing individual contracts in isolation, construction organizations must become outcome-led, aligning owners, engineers and contractors around shared objectives and value creation across the asset lifecycle. To do that, the delivery model itself also needs to evolve.

Today’s models remain largely engineering-centric, with responsibility for digital innovation often absorbed into traditional roles.
Instead, digital transformation integrators should have an explicit place alongside engineering and delivery partners, bringing dedicated AI, data and digital expertise into the project from the outset.

The role of these digital transformation integrators is not simply to implement technology, but to continuously challenge and evolve how projects are designed, delivered and operated. This ensures reinvention and innovation have clear ownership rather than relying on engineering teams to drive that agenda alongside their core responsibilities.

This new operating model creates the conditions for delivering more projects, more efficiently and with greater certainty. But achieving those outcomes also requires changing how the work itself gets done.

Updating the operating model:
Key actions

Embed desired outcomes from the earliest stages of planning.

Align commercial incentives around long-term asset performance rather than individual project milestones.

Make decisions based on total cost of ownership rather than the cost of delivery alone.

Establish the digital integrator as an explicit role in the delivery model, bringing the dedicated AI, data and digital capabilities needed to challenge traditional engineering-centric approaches and transform project and asset performance.

✓

Change 2: Ways of working

An outcome-led operating model calls for new ways of working. As AI capabilities mature, the opportunity is no longer simply to automate existing processes, but to redesign them around people and AI agents working together from the outset.

Core tasks within design development, engineering reviews, cost estimation, procurement planning and commissioning can all be reimagined around collaborative human-AI workflows that improve quality, reduce rework and accelerate delivery.

For example, AI can continuously review evolving designs for constructability, cost, schedule and compliance, providing engineers with immediate feedback rather than waiting for formal review gates. This allows teams to identify issues earlier, make decisions faster and reduce the costly rework that often occurs later in construction.

But greater AI autonomy also requires a fundamental shift in governance. Policies, controls and oversight need to be embedded into these new ways of working so that organizations can define what agents can do, where human intervention is required and how risk is managed. An agentic control plane is needed to provide that governance layer, allowing organizations to increase autonomy while maintaining appropriate control and accountability.

Human-AI workforce: Key actions

Redesign processes around human-AI collaboration rather than automating existing workflows.

Shift decision-making earlier in the project lifecycle to reduce downstream rework and risk.

Evolve governance for an agentic environment, using an agentic control plane to embed the policies, controls and oversight that govern how AI agents operate as autonomy increases.

✓

Change 3: Workforce capabilities

Redesigning how work gets done also means rethinking the skills and expertise needed to deliver it. Existing workforce shortages, now compounded by surging demand for data center development, make it increasingly important to amplify the expertise of engineers, estimators, quantity surveyors and project managers.

This is where AI can play a useful role. For example, intelligent tools can review thousands of pages of engineering documentation, compare design revisions or generate quantity take-offs in minutes rather than days. Engineers remain responsible for validating decisions, but AI frees them to focus more of their time on solving complex technical problems.

AI can also help organizations capture and transfer specialist knowledge across projects and sectors. For example, proven use cases developed in oil and gas can inform the delivery of data centers and other energy infrastructure, helping construction teams apply established engineering and project delivery practices in their industry.

Realizing this opportunity requires new competencies. Engineers, project managers and commercial professionals will increasingly need to understand not only how to use AI tools, but how to work alongside them—deciding when to rely on AI, when to challenge its recommendations and how to adapt their day-to-day work to focus on higher-value tasks.

Building workforce capabilities:
Key actions

Invest in AI literacy across engineering, commercial and project delivery teams.

Redesign roles so people focus on high-value tasks.

Capture and transfer specialist knowledge from related industries using AI, helping teams apply proven expertise to infrastructure challenges.

Embed human oversight into AI-enabled workflows to ensure engineering quality, safety and accountability.

✓

Change 4: Partnership ecosystem

Even the most capable workforce cannot deliver complex infrastructure programs in isolation. Most major projects rely on complex networks of owners, designers, contractors, suppliers and technology partners. Yet many still operate through fragmented contracts, disconnected systems and sequential handoffs that limit visibility, slow decision-making and create unnecessary rework.

Delivering construction programs at scale, including the energy transition, requires a more connected partner ecosystem. Rather than optimizing individual organizations, successful projects will increasingly optimize outcomes across the entire value chain. That means sharing information earlier, collaborating continuously and working from common digital platforms that give every participant access to the same trusted data and project insights.

The industry is already beginning to move in this direction. Early Contractor Involvement (ECI) is becoming increasingly common, bringing contractors into projects sooner so constructability, cost and delivery risks can be addressed before designs are finalized. At the same time, Modern Methods of Construction (MMC) and standardized product platforms—such as the UK's TIP 2030 Product Platform Rulebook—are encouraging repeatable designs, industrialized delivery and greater collaboration across the supply chain. The rapid growth of offsite construction also reflects this broader shift toward more integrated ways of delivering infrastructure.

AI has the potential to accelerate these models further by connecting people, data and decisions across organizational boundaries. Rather than simply improving how individual organizations perform, it enables the entire delivery ecosystem to operate as a more coordinated, intelligent network.

Driving collaboration: Key actions

Shift to an alliance-based model that brings owners, designers, contractors and technology partners together earlier.

Replace fragmented handoffs with shared digital environments that provide every partner with access to trusted information.

Align commercial incentives around program outcomes rather than individual contractual performance.

Standardize delivery through MMC, reusable product platforms, and industrialized approaches that improve quality, repeatability, and speed.

✓

Change 5: The technology foundation

Connecting people, workflows and partners requires a technology foundation designed around how infrastructure will be delivered and operated, not simply the systems used to manage individual projects today. Yet many construction organizations still rely on fragmented legacy systems built to manage documents, transactions and project activities in isolation.

Cloud adoption is helping modernize the industry’s technology estate, with 60% of construction firms expected to adopt cloud-based construction management systems in 2025.⁷ But modernization needs to go further. Organizations must reduce technical debt and evolve existing systems to support AI, while also building the new enterprise technology stack required for an agentic future—one capable of supporting AI agents, greater autonomy and new forms of human-AI collaboration.

That shift also requires organizations to rethink how work gets done and the technology that supports it. Core enterprise processes designed around traditional roles and workflows cannot simply have AI layered on top. They will need to be reinvented around a hybrid human-AI workforce, with interoperable platforms and a unified digital backbone connecting engineering, commercial and operational data across the asset lifecycle.

The technology foundation should therefore be defined with the future operating model in mind, beginning as early as the final investment decision (FID). By working backward from how an asset will ultimately be delivered, operated and maintained, organizations can determine the data, systems, agentic capabilities and interoperability requirements needed across design, construction, commissioning and operations—and build toward that future rather than trying to connect everything after delivery.

Modernizing the tech foundation:
Key actions

Define the future operating model at FID and work backward to establish the data and technology foundations required across the asset lifecycle.

Modernize the existing technology stack to reduce technical debt and create an AI-ready foundation that can support new capabilities as they emerge.

Reinvent core enterprise processes for a hybrid human-AI workforce, rather than layering AI onto processes designed around traditional ways of working.

✓

⁷ https://scoop.market.us/construction-software-statistics/

Where AI changes the game

If organizational transformation creates the conditions for change, AI becomes the engine that accelerates it. Its greatest value comes when it is embedded across the entire capital delivery lifecycle rather than deployed as isolated point solutions. Here we explore high-value use cases across the construction lifecycle.

Plan: The biggest gains come before construction starts

AI can assess site suitability, identify environmental and regulatory risks, detect design clashes and support greater standardization. But the bigger opportunity is to improve the decisions made before construction begins. By connecting data, decisions and workflows, AI-powered feedback loops can help teams anticipate emerging risks, model forward-looking scenarios and rapidly evaluate trade-offs across cost, schedule, safety and performance. With appropriate risk controls and human oversight, teams can understand the implications of different choices earlier and act with greater speed and confidence, avoiding unnecessary pauses and costly decisions that become difficult to reverse once construction begins.

The potential impact of this shift is significant. Our analysis reveals that AI-enabled approaches can reduce engineering, design and delivery effort by 30–50% and help infrastructure programs move 20–30% faster. However, the value isn’t simply fewer hours or lower costs. By accelerating design and decision-making, organizations can free up scarce engineering capacity, keep projects moving and address risks while there is still time to act—allowing gains made during planning to carry through construction and commissioning.

Build: Enable safer, more efficient delivery

During construction, AI provides real-time visibility into progress, quality, materials and emerging risks while improving reporting and decision-making. It also enhances site safety through predictive analytics and computer vision, with some applications reducing safety incidents by 36%.⁸

Operate: Designing for tomorrow

As control rooms and operational environments become increasingly autonomous, today's infrastructure must be built with tomorrow's operations in mind. Digitally connected, AI-ready assets will be better positioned to support more intelligent, efficient and sustainable operations throughout their lifecycle.

Maintain: From reactive to predictive

AI transforms asset maintenance by analyzing real-time sensor data to detect anomalies, predict equipment failures and recommend timely interventions. This shift from reactive to proactive maintenance improves safety, reliability and cost efficiency, reducing equipment downtime by up to 50%.⁹

⁸ https://www.sitesafetyai.com/blog/construction-safety-statistics-2023
⁹ https://artesis.com/10-proven-machine-maintenance-strategies-that-cut-downtime-in-2026/#key-benefits-for-reducing-downtime

Capital project excellence in Offshore Wind

Capgemini has published a new research report, “Capital Project Excellence in Offshore Wind,” examining why projects continue to face cost overruns, delays, and margin pressure despite advances in turbine technology and installation methods. Based on a survey of 150 senior offshore wind professionals across Europe, the report finds that only 13% of respondents rate recent projects as performing well on cost, while 92% have observed cascading rework caused by late design freeze.

More importantly the study reveals the industry’s biggest capability gap: 57% of executives say decision speed and quality will be critical to future success, yet only 17% believe their organization performs well today. It explains how integrated digital and AI-enabled project controls can help close this 40-points gap by providing earlier risk detection, faster impact assessment, and decision-ready options across planning, costs, risk, contracts and interfaces.

Download full report

Seizing the opportunity ahead

For decades, the construction industry has trailed other sectors in adopting advanced technology. While that has contributed to today's delivery challenge, it also presents a unique opportunity. Rather than following the same incremental modernization journey as other sectors, construction has the chance to leapfrog it, embracing AI-native ways of working while continuing its broader digital transformation.

Getting there will require more than applying AI to traditional ways of working. Construction’s engineering-led delivery models have evolved largely through incremental innovation, but the shift to an AI-enabled industry demands a more fundamental transformation. Digital capabilities need to become integral to how projects are structured and delivered, alongside engineering expertise, with new operating models, technology foundations and ways of working designed around what AI now makes possible.

For organizations willing to rethink how they operate, the rewards extend well beyond productivity. AI-enabled delivery models can shorten project timelines, reduce costs, improve margins and increase delivery capacity at a time when demand for energy infrastructure has never been greater. As project owners increasingly seek partners that can deliver faster, more efficiently and with greater certainty, the ability to combine AI with new operating models, technology platforms and ways of working will become a defining source of competitive advantage.

Authors

Paul Haggerty

Vice President, UK Market Head of Energy Transition & Utilities, Capgemini

Claire Gauthier

Executive Vice President, Global Lead of Energy & Utilities, Capgemini