Decoding HFs: what incident data reveals through SHIELD analysis

This presentation explored how structured human factors (HF) taxonomies, specifically the SHIELD framework, can enhance safety learning from maritime incidents. Drawing on a growing database of accident investigation reports, the talk demonstrated how SHIELD enables systematic classification and causal weighting of human and organisational factors. A detailed case study of the EVER FORWARD grounding incident illustrated the use of SHIELD in identifying performance breakdowns across cognitive, communicative, and operational dimensions.

Additionally, the presentation highlighted the benefits of integrating success-based learning to supplement incident-driven data, thus broadening the scope of safety insights. This approach aligns with emerging trends in resilience engineering and proactive safety management. The session concluded with a discussion on future directions, including AI-supported HF data processing using Large Language Models to automate analysis and scale research impact across the maritime domain.