AI for industrial safety: A comprehensive mapping of the opportunities and risks of deployment
TL;DR
This document presents an overview of the opportunities and concerns associated with AI deployment through a graphical representation of six application domains, depicted as the petals of a flower: anticipating risks; distancing operators from the most hazardous situations; protecting exposed operators; enhancing human knowledge and decision-making; automating the monitoring and maintenance of technical systems; and, finally, enriching experience feedback.
Abstract
Artificial Intelligence (AI) is rapidly making its way into all areas of corporate life, including the field of safety. Its development is generating high expectations, concerns, and questions regarding its deployment, governance, and organizational impact. This document presents an overview of the opportunities and concerns associated with AI deployment through a graphical representation of six application domains, depicted as the petals of a flower: anticipating risks; distancing operators from the most hazardous situations; protecting exposed operators; enhancing human knowledge and decision-making; automating the monitoring and maintenance of technical systems; and, finally, enriching experience feedback. As part of the FonCSI initiative "Safety of the Future: “Living with” Uncertainty, Complexity, and New Expectations", this document draws on the academic and technical literature as well as on the work of a foresight workshop that has brought together, since 2025, around twenty experts from industry and regulatory authorities, along with external contributors.