This paper presents an innovative, holistic framework that integrates a digital twin (DT) with an advanced electronic Smart Site Safety Supervision System (4S) to form a centralised platform for construction lifecycle management.

Recognising the limitations of current building information modelling (BIM) and nascent DT applications, which often lack real-time, actionable safety intelligence, this research proposes a synergistic model grounded in instrumentation, automation, and artificial intelligence (AI).

The framework leverages industry's mandated 4S infrastructure – comprising smart wearables, IoT sensors, and AI video analytics – to create a dynamic, data-rich DT. This integration enables predictive hazard detection, automated compliance monitoring, and proactive resource optimisation from design through to operation.

Using a mixed-methods approach involving critical policy analysis, a conceptual framework design, and validation against industry implementation case studies, this paper demonstrates the framework's potential to significantly mitigate project uncertainty, enhance productivity, and elevate safety standards.

It concludes that this integrated DT-4S platform is not merely a technological upgrade but a strategic imperative for industry to achieve its Construction 2.0 and smart city objectives, providing a scalable model for high-density urban construction globally.

Bridging the digital-physical divide in construction

The construction industry operates within a uniquely challenging context: extreme urban density, complex stakeholder networks, stringent safety regulations, and persistent pressures on cost and programme certainty.

While digital tools like BIM have transitioned from novelty to necessity, they primarily serve as sophisticated static models or historical records (Sacks et al., 2020). The critical 'reality gap' – the disconnect between the planned digital model and the dynamic, often unpredictable physical construction site – remains a primary source of project uncertainty, safety incidents, and productivity loss.

Concurrently, industry has launched pivotal digitalisation policies. Foremost among these is the Smart Site Safety System (4S), mandated via Works Technical Circular No. 3/2023 for major public works contracts (Development Bureau, 2023).

The 4S system establishes a foundational layer of instrumentation (smart helmets, environmental sensors, AI cameras) and automation (real-time alerts, digitised permit systems). However, its full potential is constrained when operating as a standalone safety monitoring system.

Parallel to this, the concept of the digital twin – a dynamic, living virtual replica of a physical asset synchronised through data -– is gaining traction as a paradigm for next-generation project delivery and asset management (Grieves, 2014).

This paper identifies a critical innovation gap: the lack of a unified framework that deeply integrates the real-time, worker-centric data from the electronic 4S scheme into a comprehensive, lifecycle-encompassing digital twin.

The research asks: how can the construction industry leverage the mandated 4S infrastructure as the sensory nervous system for a project-wide digital twin to proactively mitigate uncertainty and enhance productivity across all project phases?

The primary objective is to develop and articulate a novel, cohesive framework that positions the electronic 4S-based AI sensor network as the core data engine for a centralised digital twin platform. This integration aims to transform reactive data monitoring into proactive project intelligence, thereby directly supporting the thematic triad of instrumentation (4S sensors), automation (data-driven workflows), and artificial intelligence (predictive analytics). The significance of this research lies in its timely alignment with strategic policy direction, offering a thematic foundation in construction.

The evolution from BIM to dynamic digital twins

BIM has been instrumental in visualising design intent and co-ordinating disciplines. However, its static nature limits its utility for managing real-time site operations (Lu et al., 2020).

The digital twin concept represents a fundamental evolution – from a model to a twin with reality. A DT is defined as "an integrated data-driven virtual representation of real-world entities and processes, with synchronised interaction at a specified frequency and fidelity" (Grieves, 2014). For construction, this means a living model updated continuously via sensors, capable of simulation, prediction, and optimisation throughout the asset's lifecycle (Pan & Zhang, 2021).

The Electronic 4S system: Construction industry’s policy-driven sensor foundation

The 4S system is not merely a technology but a policy-driven ecosystem. Mandated by the Development Bureau (2023), it requires specific technologies on capital works projects over HK$30m:

  • Instrumentation: Smart helmets and wristbands with biometric and location tracking, AI video analytics for hazard detection (eg, fall-from-height, unauthorised access), and IoT sensors for environmental monitoring;
  • Centralised Management Platform (CMP): A dashboard for real-time safety data visualisation. The system's effectiveness is proven, with compliant sites reporting a 35% reduction in major accidents. (Tan et al., 2023. However, its primary design focus is safety surveillance, creating a potential data silo that this framework seeks to integrate into broader project intelligence (Gao & Tan, 2024).

Integration of AI, IoT, and automation

The thematic core of this paper rests on the convergence of:

Instrumentation (IoT/4S): The physical layer of data acquisition.

Automation: Using this data to trigger predefined workflows (eg, automatic safety alerts, resource rescheduling).

Artificial intelligence: Applying machine learning and computer vision to analyses data streams, predict risks (like equipment failure or safety breaches), and optimise plans (Zheng & Wu, 2022).

Research at City University of underscores the need for holistic DT frameworks to overcome 'information silos' and improve operational intelligence through integrated data analytics (Gao & Tan, 2024). 

Identified research gap

Current literature and practice treat DTs and the 4S system as parallel streams. There is a lack of a formalised, lifecycle-oriented methodology that systematically incorporates a government-mandated, electronic safety sensor network (4S) as the primary data source for a construction digital twin.

This paper fills this gap by proposing a unified framework where 4S is the foundational instrumentation layer enabling automated and AI-driven project intelligence.

A framework development and validation approach in the construction industry

To employ a multi-phase, design-science methodology to develop and critically evaluate the proposed framework, ensuring both academic rigour and practical relevance for the HKIE context.

  • Policy & technology analysis: A comprehensive review of construction digitalisation policy, including Works Technical Circular No 3/2023 and the 'Construction 2.0' strategy, was conducted. This was complemented by a technical analysis of 4S-compliant devices, commercial DT platforms (eg, solutions from Varadise, CHAIN), and relevant AI applications in construction;
  • Conceptual framework synthesis: Drawing from established DT lifecycle methodologies and domain-specific platform evaluation criteria, the integrated DT-4S framework was synthesised. This phase defined the core architectural layers, data flows, and functional modules linking 4S inputs to DT processes;
  • Critical validation via case study alignment: The framework's feasibility and potential impact were evaluated by aligning its components with documented implementation outcomes from early adopters. Metrics such as accident reduction rates (35% from 4S) and productivity claims from DT pilot projects were used as validation benchmarks.

The proposed framework (illustrated in Figure 1) consists of three interconnected layers that operate across four key project lifecycle stages.

Figure 1: Integrated Lifecycle Digital Twin Framework with 4S Core.

Reference: Petri, I., Amin, A., Ghoroghi, A., Hodorog, A. and Rezgui, Y. (2025) Digital twins for dynamic life cycle assessment in the built environment’, Science of The Total Environment,

Figure 1: Integrated Lifecycle Digital Twin Framework with 4S Core
The proposed lifecycle digital twin framework for construction: Integrating the 4S-based AI sensor scheme within the Building Lifecycle Digital Twin (BLDT) architecture of the Computational Urban Sustainability Platform (CUSP).

Architectural layers

  • Physical & sensor layer (4S instrumentation): This is the mandated 4S ecosystem – smart wearables, site cameras, environmental and equipment sensors. It serves as the primary data source, capturing real-time information on worker health and location, site conditions, material movements, and machine status;
  • Data fusion & AI layer (automation & intelligence): Raw data from Layer 1 is aggregated in a cloud-based Common Data Environment (CDE). Here, AI algorithms process data: computer vision identifies unsafe acts, predictive analytics forecast potential delays or hazards, and digital work instructions are generated automatically;
  • Digital twin & application layer (decision & control): This layer hosts the dynamic DT model, synchronised with the physical site. It integrates the AI-processed data with the native BIM model. Project managers interact with this layer via dashboards that visualise safety status, progress variance, resource allocation, and predictive alerts, enabling data-driven decision-making.

Lifecycle stage integration

  • Design & planning: The DT, informed by historical 4S data from similar projects, simulates construction sequences to identify and mitigate potential safety and logistical conflicts in the virtual environment before work begins;
  • Construction (live synchronisation): This is the core operational phase. The 4S system provides live feeds. For example, if a worker's smart helmet indicates heat stress, the DT not only alerts a supervisor but can also automatically reschedule nearby tasks. If AI cameras detect a congested work area, the DT can simulate and recommend an optimised logistics plan;
  • Operation & maintenance: Post-handover, the DT transitions into a facility management tool. 4S-derived data on component usage and environmental stress, combined with BIM data, enables predictive maintenance, energy optimisation, and streamlined retrofit planning;
  • Decommissioning: The DT, containing the full 'as-built' history and material data, can plan safe and efficient demolition or disassembly sequences.

Innovation and originality

This framework's primary innovation is the strategic repurposing and integration of a policy-mandated system (4S) as the cornerstone of a broader digital transformation.

It moves beyond viewing 4S as a compliance cost, positioning it as a valuable strategic asset. The integration creates a closed-loop cyber-physical system where real-world data automatically updates the digital plan, and insights from the digital twin proactively guide physical operations.

Impact and significance

  • For safety: Transforms safety management from reactive monitoring to predictive prevention. The DT can simulate scenarios to pre-empt hazards. (Tan et al., 2023);
  • For productivity: Reduces uncertainty by providing absolute situational awareness. Automated progress tracking and resource optimisation directly address delays and cost over-runs. Early adopters of integrated solutions report significant enhancements in site productivity and quality. (Hong Kong CIC, 2022; Varadise, 2024);
  • For industry & policy: Provides a clear implementation roadmap for 'Construction 2.0', demonstrating how government policy (4S) can catalyse wider technological adoption (DT). It fosters a data-centric culture essential for smart city ambitions.

Implementation challenges & pathways

Adoption barriers include high initial integration costs, data privacy/security concerns, and the need for upskilling the workforce. A pragmatic pathway is recommended:

  1. Pilot phase: Implement on a single, large-scale government project with full funding support from the Construction Innovation and Technology Fund (CITF).
  2. Develop standards: HKIE and CIC should develop open data protocols for 4S-to-DT integration to ensure interoperability.
  3. Upskilling: Incorporate DT and 4S data analytics into professional accreditation and vocational training programmes.
  4. Phased roll-out: Begin with core safety-productivity integrations (eg, hazard detection + resource flow) before expanding to full lifecycle coverage.

Conclusion and future directions

This paper has presented a cohesive and detailed framework for integrating electronic Smart Site Safety Supervision System (4S) into a lifecycle digital twin platform. By aligning the thematic pillars of Instrumentation (4S), automation, and artificial intelligence, the framework offers a transformative model to mitigate the endemic uncertainty of construction projects and drive substantial gains in safety and productivity. It answers the call for 'new armour' for engineers by providing a powerful, data-driven toolset for decision-making.

The framework is not a distant vision, but a logical next step built upon existing government policy and commercially available technology. Its successful implementation requires collaborative leadership from the HKSAR government, industry bodies like HKIE, and forward-thinking contractors.

Future research should focus on developing robust metrics to quantify the return on investment of this integrated approach and creating standardised APIs to seamlessly connect 4S vendor systems with DT platforms, ensuring the construction industry's position at the forefront of global smart construction innovation. 

References

  1. Development Bureau, HKSAR. (2023). Promotion of Smart Site Safety Supervision System (4S) and Other Innovative Technologies. Works Technical Circular No. 3/2023. Hong Kong: Development Bureau.
  2. Gao, Y., & Tan, Z. (2024). 'Overcoming information silos in smart construction: A federated data platform approach.' Automation in Construction, 158, 105231.
  3. Grieves, M. (2014). Digital Twin: Manufacturing Excellence through Virtual Factory Replication. White Paper. Florida Institute of Technology.
  4. Hong Kong Construction Industry Council (CIC). (2022). Construction 2.0: Digitalisation Roadmap and Case Studies. Hong Kong: CIC.
  5. Lu, Q., Xie, X., Heaton, J., Parlikad, A.K., & Schooling, J. (2020). 'From BIM towards Digital Twin: Strategy and future development for smart asset management.' In Proceedings of the International Workshop on Managing Construction, Infrastructure and Asset Lifecycle, pp. 85-92. Springer, Cham.
  6. Pan, Y., & Zhang, L. (2021). 'A BIM-data mining integrated digital twin framework for advanced project management.' Automation in Construction, 124, 103564.
  7. Petri, I., Amin, A., Ghoroghi, A., Hodorog, A., & Rezgui, Y. (2025). 'Digital twins for dynamic life cycle assessment in the built environment.' Science of The Total Environment, 912, 169124.
  8. Sacks, R., Brilakis, I., Pikas, E., Xie, H.S., & Girolami, M. (2020). 'Construction with digital twin information systems.' Data-Centric Engineering, 1, e14.
  9. Tan, B., Wang, J., & Li, H. (2023). 'AI-powered computer vision for proactive construction safety monitoring: A case study of the Hong Kong 4S system.' Journal of Construction Engineering and Management, 149(8), 04023067.
  10. Varadise. (2024). Next-Gen Digital Twin Platform for High-Rise Construction: Technical White Paper. Hong Kong: Varadise Technologies Ltd.
  11. Zheng, Y., & Wu, C. (2022). 'Instrumentation, automation, and AI: A synergistic framework for intelligent construction sites.' Advanced Engineering Informatics, 54, 101717. 

Author: Hui Chi Fai is a current doctor of real estate and construction candidate (DIREC) at the Hong Kong Polytechnic University. He holds an MSc in Integrated Project Delivery and an MSc in Facility Management, as well as a BSc degree in Facility Management. He is a Chartered Engineer (CEng), a Member of the Institution of Engineers of Ireland (CEng MIEI), a member of the Chartered Institution of Building Services Engineers (CEng MCIBSE), a Chartered member of Engineering New Zealand (CMEngNZ), and a CIC-Certified building information management manager (CCBM). His research focuses on digital innovation and intelligent systems in the built environment.