Pupillometry based real-time monitoring of operator's cognitive workload to prevent human error during abnormal situations

Abstract

Ensuring process safety is widely regarded as critical to sustainable development. Despite numerous initiatives to improve process safety, severe accidents continue to occur in the process industries. This is commonly attributed to increased complexity of plants, reduced staffing levels, and the consequent cognitive challenges faced by process operators. Statistics indicate that human error is a dominant contributor in over 70% of accidents. Human error is traditionally considered only in the process design stage. In this work, we propose a methodology based on pupillometry (the measurement of pupil diameter) to noninvasively estimate the cognitive workload of control room operators in real time during process operations. Experimental studies conducted on 44 participants reveal that changes in the operators’ pupil size during abnormal situations contain distinct signatures of their ability to successfully manage the process abnormality. Our results demonstrate that pupillometry has the potential for online performance monitoring of control room operators.