Digital twins are moving from design support tools to strategic platforms for whole-life asset intelligence. Their evolution reflects a broader shift across the digital construction sector: from capturing information to understanding risk, anticipating future conditions and improving decision-making at every stage of a programme. Speaking with Simon Foulds, Nathan Marsh, Regional Executive and Senior Vice President for EMEA at Bentley Systems, outlines how digital twins can underpin a new model of asset lifecycle risk management, and what must change for predictive analytics to gain full industry confidence.
From static models to living digital twins
Digital twins initially gained traction through building information modelling (BIM) and the transition from 2D to 3D environments. Marsh argues that this early focus, while necessary, barely scratched the surface of what digital twins in construction can achieve today. Instead of static representations, modern infrastructure digital twins are becoming ‘living’ systems capable of analysing engineering, cost, schedule and risk data in one environment.
For Marsh, the value lies in ‘foresight, predictability and risk reduction. A digital twin designed with this purpose can ingest a wide range of information, from engineering drawings and construction schedules to supply-chain inputs and operational performance metrics. By consolidating data that would otherwise sit in fragmented systems, digital twins offer a continuous line of sight from design and engineering through to construction and operations.
Central to this is the ability to present the right information at the right time. Marsh highlights the importance of unified dashboards that scale with increasing data volumes. For engineers, contractors and asset owners, this creates a single point of truth, enabling scenario modelling, optioneering and early risk identification. The outcome is a platform capable of supporting complex major projects by helping teams anticipate issues before they escalate.
The trust gap in predictive analytics
While predictive analytics in construction are widely recognised as transformative, their adoption remains constrained by one persistent barrier: trust. Marsh emphasises that project teams may not fully trust their own data, let alone datasets provided by partners or competitors. Fragmentation across project phases and organisations compounds these concerns.
To realise the potential of predictive analytics and machine-learning-enabled insights, Marsh argues for stronger foundations in data provenance, data governance and structural consistency. Understanding who created the data, when, where and under which conditions is critical, particularly as AI models depend on clarity around source data and code.
A controlled and auditable data environment is essential. This includes secure data management, clear ownership, consistent formatting and transparent access rights. For investors and asset owners, confidence increases when digital assets follow robust information security practices and structured governance frameworks.
Standards and contractual obligations
Achieving this level of control depends heavily on the adoption of shared standards. Marsh sees a clear role for both principles-based and, in some sectors, prescriptive standards. Highly regulated areas such as nuclear, defence or energy infrastructure may require government or regulator involvement, while ISO and BSI standards remain key references across the industry.
Major programmes, often spanning 15 to 20 years, can define their own data frameworks in collaboration with stakeholders, investors and programme partners. Marsh describes the need for both ‘carrot and stick’: strong incentives for consistent data practices, but also contractual obligations ensuring compliance. Without clear expectations embedded at the procurement stage, data quality will remain variable and confidence in predictive tools will continue to lag.
Filtering signal from noise in real-time monitoring
Real-time performance monitoring is producing more data than ever. Yet without the ability to distinguish signal from noise, the volume of information can hinder rather than support decision-making.
Here, Marsh returns to the central role of the digital twin. When tuned and trained effectively, a digital twin in infrastructure projects filters what matters: safety indicators, schedule dependencies, cost pressures and other mission-critical triggers. Personalisation of dashboards and alerts means each member of a programme team sees data aligned to their decision-making needs.
Preparing for decision cycles, whether three, six or nine months ahead, is essential. Marsh recommends tiering data according to urgency, certainty and analytical depth. Some decisions require rapid, high-level information; others need time for more detailed analysis. Adaptive thresholds help track deviations from baselines and highlight changes early, supporting proactive risk management rather than reactive problem-solving.
Shared accountability for digital risk forecasting
As digital tools become part of everyday governance, questions arise over accountability when forecasts fail to match real-world performance. Marsh is clear: accountability is shared. Asset owners and investors must set the tone through a clear data strategy, defined risk appetite and governance frameworks. Operators need to articulate how data requirements evolve over time. Technology providers, including Bentley Systems, contribute sector knowledge built over decades.
This shared model mirrors the broader nature of major infrastructure projects: complex ecosystems that rely on collaboration, transparency and consistent standards. Marsh notes that openness is core to Bentley’s approach, ensuring applications can integrate across supplier networks while maintaining structured and reliable data environments.
Towards predictable, transparent and safe major projects
Digital twins are shifting from design tools to platforms that support risk visibility, operational foresight and strategic decision-making across the entire asset lifecycle. Their value depends on trusted data, robust governance, shared standards and alignment between owners, operators and technology providers. With predictive analytics, real-time monitoring and integrated dashboards, digital twins can help major programmes identify issues early, reduce uncertainty and deliver safer, more predictable outcomes.
As Marsh highlights, the path forward requires both direction and discipline. With consistent standards, transparent data structures and a commitment to open collaboration, digital twins can deliver the clarity and control that modern infrastructure demands.
Read next: High-volume night works on the M4 – Build in Digital
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
