DTSafe-O

Despite the digitalization of architecture, engineering, and construction, construction-site safety is still limited by fragmented data, non-computable regulations, and slow decision processes. This paper presents DTSafe-O, a semantic digital twin for safety management that integrates an ontology formalized in Web Ontology Language 2 (Description Logic profile), guided by Semantic Web Rule Language and Shapes Constraint Language and connecting Internet of Things sensors, building information modeling, and regulatory documents. The contribution is twofold: (a) a computable formalization of safety-critical requirements enabling explainable and verifiable inferences and (b) a validation pipeline combining reasoning, compliance checking, and performance metrics, supported by replicable materials in a public repository. Results show that merging digital twins with semantic technologies improves consistency, timeliness, and auditability of safety decisions, laying the ground for scalable industrial adoption. The study also discusses limitations and opportunities for extending the approach to infrastructure domains and advancing spatiotemporal reasoning and uncertainty management.