Data-Driven Construction Safety Information Sharing System Based on Linked Data, Ontologies, and Knowledge Graph...

Area of Science:

• Construction Safety Management
• Data Science Applications
• Industry 4.0

Background:

• High accident, injury, and fatality rates persist in the construction industry.
• Existing accident data is often inconsistently formatted, non-machine-readable, and inaccessible, limiting learning opportunities.
• Advances in data science and Industry 4.0 offer potential for improved safety information sharing.

Purpose of the Study:

• To address challenges in construction safety data accessibility and reusability.
• To propose a novel information sharing system leveraging linked data, ontologies, and knowledge graphs.
• To enable proactive accident prevention through enhanced data utilization.

Main Methods:

• Developed an ontological approach to semantically model safety information and formalize accident case knowledge.
• Implemented a multi-algorithmic approach for automatic conversion of accident data to Resource Description Framework (RDF).
• Utilized the SPARQL protocol for enabling efficient query functionalities on the structured data.

Main Results:

• Confirmed the effectiveness and efficiency of the proposed system through trials on 200 real accident cases.
• Demonstrated significant improvements in information access, retrieval, and reusability.
• Validated the system's capability to capture and disseminate insights from past mishaps.

Conclusions:

• The developed system facilitates a new "open" information sharing paradigm for construction safety.
• This approach has major implications for Industry 4.0 and data-driven safety management applications.
• The system enhances the potential for learning from past accidents to proactively prevent future incidents.