Safety decisions are often made under pressure, with incomplete information and changing conditions. This session examined how AI can help close that gap by bringing together data that would otherwise remain fragmented across systems, teams, and timeframes.
The discussion focused on practical HSE use cases, from predicting equipment failures and environmental incidents to supporting faster response during emergencies. By combining sensor data, maintenance logs, environmental conditions, and operational history, AI can surface patterns that are difficult to detect manually, enabling earlier intervention and more informed choices.
A significant part of the conversation centred on people rather than technology. Introducing AI into hazardous environments raises legitimate concerns around surveillance, accountability, and trust. Panellists stressed that AI must be deployed with transparency, clear intent, and strong collaboration with the workforce. Ensuring that workers can question, challenge, or override AI recommendations was seen as essential to maintaining safety and confidence.
Looking ahead, the session explored how explainable and auditable AI models could strengthen HSE governance. From neuro symbolic approaches to shared safety knowledge frameworks, the panel agreed that the future of AI in HSE lies in systems that support human judgement rather than attempt to replace it.
Speakers
- Grace Chan , Venture Principal at bp Ventures
- Dr. Reza Khaninezhad , Data Science Manager, Apache
- Riad Efendi , QHSE and Digital Solutions Leader, TechnipFMC
