Digital twins can deliver large productivity and sustainability gains in advanced construction pilots but those wins are not yet industry‑wide; scaling value will depend on interoperability standards, new contracting and data‑governance arrangements, and investment in skills and delivery models.

Main Finding

Digital twin (DT) technology can materially improve construction lifecycle performance beyond Building Information Modelling (BIM), but widespread industry benefits are limited by interoperability, data-quality, governance, and organisational barriers. DTs are a socio-technical transformation requiring standards, collaborative delivery models, and workforce capability building to realise scalable economic value.

Key Points

• Scope of review: 160 peer‑reviewed studies (2018–2026) selected from 463 Scopus records via a PRISMA-guided systematic review; inter-rater reliability Cohen’s κ = 0.83.
• How DTs extend BIM (three core differences):
• Bidirectional, automated physical↔digital data exchange (not just static models).
• Integration of heterogeneous, real‑time sources (IoT sensors, operational systems, etc.).
• Lifecycle continuity from design through operation and end‑of‑life, preserving data across handovers.
• Reported performance gains (from advanced pilots; not industry averages):
  • Rework and logistics reductions up to ~80%
  • Cost savings ~5%
  • Schedule acceleration ~2 months
  • Energy reductions 15–30%
  • Maintenance cost reductions 10–25% Note: these are case‑level results from high‑performing implementations and should not be treated as typical benchmarks.
• Principal barriers to adoption:
  • Interoperability gaps and lack of standards
  • Data quality and continuity problems at handover
  • Low digital maturity and uneven capabilities across supply chains
  • Misaligned stakeholder incentives and fragmented project delivery models
  • Paper‑based or legacy regulatory/compliance processes
• Framing: DT adoption is socio‑technical — requires governance, standards, procurement change, and workforce development, not just technology deployment.
• Recommended future research: scalable interoperability solutions, longitudinal lifecycle value validation, human‑centred adoption strategies, and sustainability assessment methods.

Data & Methods

• Data sources: Scopus database search yielding 463 records (2018–2026).
• Selection process: PRISMA-guided screening and eligibility filtering; final sample 160 peer‑reviewed studies.
• Quality control: Inter‑rater reliability test with Cohen’s κ = 0.83 (substantial agreement).
• Evidence types: mix of conceptual papers, case studies, pilot deployments, and limited larger empirical evaluations; advanced implementations provide most of the quantitative performance figures.
• Limitations: evidence skewed toward pilot/high‑performer contexts; lack of long‑panel, multi‑project longitudinal studies assessing typical returns and scalability.

Implications for AI Economics

• New data infrastructure for AI:
  • DTs generate continuous, high‑resolution operational data (IoT telemetry, usage patterns, maintenance logs) that can substantially improve AI models for predictive maintenance, scheduling, energy optimisation, and logistics.
  • Better data continuity across lifecycle phases reduces model training friction and increases the value of historical data for forecasting and causal analysis.
• Productivity and cost dynamics:
  • Reported pilot gains suggest potential for meaningful productivity improvements and cost reductions, which could shift firm-level returns and industry productivity measures if scaled.
  • However, gains are contingent on coordinated adoption; uneven uptake may produce winner‑takes‑more dynamics among technologically advanced firms.
• Investment and returns:
  • Realising DT value requires upfront investment in sensors, integration, standards, and skills. Economic viability depends on contract structures, risk allocation, and whether gains are captured by investors, owners, contractors, or operators.
  • Public procurement and large asset owners can act as powerful demand‑pulls to de‑risk early investment and set standards.
• Market structure and incentives:
  • Misaligned incentives across designers, builders, and operators hinder data continuity; new contracting and governance models (e.g., performance‑based contracts, data‑sharing agreements) are needed to internalise lifecycle benefits.
  • Standards and open interoperability reduce lock‑in and transaction costs, widening market access and competition for AI services built on DT data.
• Policy and regulation:
  • Paper‑based regulatory environments slow diffusion; digitised compliance and standardised data schemas can accelerate adoption and enable AI‑driven oversight.
  • Policy levers: funding for standards development, incentives for early adopters, requirements for data handover in public projects, and support for workforce reskilling.
• Research & measurement needs for economics:
  • Longitudinal, multi‑project studies measuring realised vs. ex‑ante gains, distributional impacts (which firms capture benefits), and externalities (energy, emissions).
  • Cost‑benefit frameworks that incorporate upfront investments, recurrent data management costs, and the value of information for AI models.
  • Methods to assess scalability and generalisability of pilot results to average projects and different regional regulatory contexts.
• Risk considerations:
  • Data governance, privacy, and cybersecurity risks can create negative externalities and require governance frameworks that affect adoption costs and social welfare.
• Practical takeaway for AI economists:
  • DTs materially expand the data frontier for AI in construction, promising productivity and sustainability gains, but economic value depends critically on institutional arrangements (standards, contracts, governance) and measured, longitudinal evidence of scalable returns.