Presentation

· Contributors · Organizations · Search Program

LLM-powered Hazard Identification and Risk Analysis for Offshore Safety Reporting

DescriptionOffshore oil and gas operations rely on timely hazard identification and workforce participation to prevent major incidents, yet conventional safety observation reporting systems remain time-consuming and burdensome. This paper presents SURE Safety, an AI-powered mobile safety reporting application designed to improve hazard documentation and participation in offshore safety programs. Motivated by persistent challenges following the Deepwater Horizon disaster and the implementation of SEMS-II SURE programs, the application integrates multimodal large language models and retrieval-augmented generation to support photo-based hazard analysis, structured report generation, and user review. Unlike traditional paper-based or digitized reporting forms, SURE Safety aims to re-engineer the reporting workflow by automatically identifying hazards, pre-populating report fields, and allowing workers to confirm or revise AI-generated content before submission. The proposed workflow is expected to reduce report completion time from approximately 10–15 minutes to 45 seconds–3 minutes, thereby lowering cognitive and temporal barriers to reporting. The system emphasizes human-in-the-loop design, preserving worker authority and reducing risks associated with automation bias or inaccurate AI outputs. This work contributes an early example of applying modern multimodal AI to offshore safety reporting and highlights the importance of designing AI systems that support, engage, and empower frontline workers in safety-critical environments.

Contributors