Dataset of characterised construction safety risks and related treatments

Carlos Osorio Sandoval, Gordon Crick, William Collinge, Karim Farghaly, Mojgan Hadi Mosleh, Patrick Manu, Clara Cheung

Research output: Contribution to journalArticlepeer-review

Abstract

The Safety Risk Library [1] is a structured database [2] that integrates knowledge drawn from multiple sources to address the problem of information disaggregation in the construction industry. This knowledge base maps construction safety risk scenarios to treatment suggestions that help designers implement the concept of prevention through design. In the context of the Safety Risk Library, risk scenarios are characterised by six data categories based on a formalised ontology [3]. To build the first iteration of the Safety Risk Library, nine different risk scenarios were identified and mapped to relevant risk treatments in focus groups. Subsequently, the Safety Risk Library was pilot tested in six construction projects, and user feedback and input were used to expand the list of risk scenarios and treatment prompts. Additionally, public press releases reporting construction accidents were analysed to identify and characterise risk scenarios, which were then mapped to appropriate treatment suggestions and included in the Safety Risk Library. This dataset can assist construction industry stakeholders in identifying, characterising, communicating and mitigating safety risks in construction projects. It can also be integrated into building information modelling environments to assist designers to implement prevention through design.
Original languageEnglish
JournalData in Brief
Early online date4 Jun 2023
DOIs
Publication statusPublished - 7 Jun 2023

Keywords

  • Design for safety
  • Building information modelling (BIM)
  • Risk scenarios
  • Prevention through design

Research Beacons, Institutes and Platforms

  • Thomas Ashton Institute

Fingerprint

Dive into the research topics of 'Dataset of characterised construction safety risks and related treatments'. Together they form a unique fingerprint.

Cite this