HSE & Sustainability

Automation, AI Enhance Management of HSE Risks and Incidents

This paper outlines how one company uses digital technologies to manage HSE risks in project delivery, developing an artificial intelligence (AI) predictive model to predict HSE risks and incidents based on historical incident data.

Artificial intelligence and smart concept
Source: Peshkov/Getty Images

As society and the economy gravitate toward cleaner and renewable energy, the energy sector continues to diversify and increase in complexity, leading to a “never normal” business portfolio and project delivery. On top of safeguarding on-time, on-budget, and on-scope delivery, this constant state of change also warrants new ways of managing health, safety, and environment (HSE).

This paper outlines how Petronas uses digital technologies to manage HSE risks in project delivery. Since 2019, the project delivery arm of Petronas has invested in developing an artificial intelligence (AI) predictive model to predict HSE risks and incidents based on historical HSE data.

Coupled with real-time data visualization and prescription of control measures, this has greatly enabled the organization to rationalize resources and implement informed, risk-based interventions. Considerable efforts have been made to operationalize the predictive tool, which includes stakeholder management and change management, as well as work process enhancement, to incorporate automation and have end-to-end oversight on risk mitigation and intervention. This provides an HSE management framework that focuses on data-driven decision-making, which translates to targeted intervention with pace.

The digital HSE plan is built on a plan, do, check, act (PDCA) concept, whereby HSE incidents that have occurred within the organization are used as a data source for analytics to visualize trends and predict HSE risks, incidents, and underlying causes. Targeted interventions then can be put in place to address the prediction results with the aim of preventing incident occurrence. Nevertheless, data from new HSE incidents will be fed to the digital solution machine-learning model to improve prediction accuracy.

As a result of implementing this plan, the company’s project delivery arm has seen an incident reduction of more than 55% for both injurious and noninjurious incidents in 2022 compared with the preceding 3-year average. In addition, the implementation of the digital technology has enabled an improved process cycle efficiency of HSE risk management from 32% to 97%, with a reduction of total work time of 143 hours.

SPE members can download the complete paper from SPE’s Health, Safety, Environment, and Sustainability Technical Discipline page for free from 18 to 31 July.

Find paper SPE 216290 on OnePetro here.