Modern chemical industries are increasingly complex and highly integrated, comprising of multiple processes and units. These complex processes are overseen by operators through human machine interfaces (HMIs). Model predictive control (MPC) is an advanced control scheme, which is increasingly used to manage individual processes and an entire chemical plant. MPC uses a predictive model to adapt to disturbances and/or maximizes productivity, while maintaining product quality. When effective, MPC shifts operators from active control to passive monitoring and intervention. However, the complexity of MPC or poor understanding among operators often prompts them to manually override the controller, especially as it differs from familiar single-loop PID controllers (Lindscheid et al., 2016). An assessment of operator interactions with the HMI for a MPC-controlled chemical process is proposed. Operator’s performance and HMI efficiency have been assessed using cognitive measures, such as eye tracking, showing its potential to capture attention during abnormal scenarios (Kodappully et al., 2016) and to link interface design to perceived workload (Ikuma et al., 2014). However, few combine MPC and cognitive tools for a holistic view of operator-MPC interactions. The overall objective of our research is to address this gap.

Previously, we have reported a testbed for understanding operator challenges in multi-unit chemical processes equipped with P/PI controllers and a high-level MPC (Ranjan et al., 2023). Here, we use the testbed to conduct eye tracking studies to examine operators’ responses to faults and disturbances. Analyzing process data and gaze patterns during simulated tasks, we observed that as operators gain familiarity with MPC and with better understanding of system dynamics, they reduce manual overrides, relying on the MPC to maintain stability. This approach enhances our understanding of operators’ attention and comprehension when managing advanced controllers in complex systems while identifying triggers for their actions.

References

Lindscheid, C., Bremer, A., Haßkerl, D., Tatulea-Codrean, A., Engell, S., 2016. A test environment to evaluate the integration of operators in nonlinear model predictive control of chemical processes, IFAC-PapersOnLine 49, 129–134.

Kodappully, M., Srinivasan, B., Srinivasan, R., 2016. Towards predicting human error: Eye gaze analysis for identification of cognitive steps performed by control room operators, Journal of Loss Prevention in the Process Industries, 42, 35-46

Ikuma, L. H., Harvey, C., Taylor, C. F., & Handal, C. (2014). A guide for assessing control room operator performance using speed and accuracy, perceived workload, situation awareness, and eye tracking. Journal of Loss Prevention in the Process Industries, 32, 454-465.

Ranjan, R., Das, L., Kaisare, N. S., & Srinivasan, R. (2023). A testbed for studying the interactions between human operators and advanced control systems. Computers & Chemical Engineering, 178, 108377.