434866 Performance Monitoring of Control Room Operators through Eye Gaze Analysis

Thursday, November 12, 2015: 2:43 PM
Salon G (Salt Lake Marriott Downtown at City Creek)
Punitkumar Bhavsar, Electrical Engineering, Indian Institute of Technology Gandhinagar, Ahmedabad, India, Babji Srinivasan, Department Chemical Engineering, Indian Institute of Technology Gandhinagar, Ahmedabad, India and Rajagopalan Srinivasan, Department of Chemical Engineering, Indian Institute of Technology Gandhinagar, Gujarat, India

Abstract

Safety in chemical process plant has become primary focus after the accidents like three mile island(1979), Esso’s Longford refinery (1998),BP’s Texas City refinery (2005) etc. to count a few. The investigation reports reveal the control room operator error as one of the leading factor for such incidents. Studies from Mannan (2004) suggest that almost 70 % of the critical incidents occurred just because of human error in plant operation. Analysis in high risk industries indicate that human errors typically originate from failure of proper situation assessment (SA) during abnormal plant conditions (Endsley 1988). It has been shown that analysis of cognitive behavior of humans can reveal their level of situation awareness, vital information that could help prevent human errors.

Studies on eye mind hypothesis suggest that gaze movement and cognitive behavior are highly correlated. Therefore, visual attention from eye gaze analysis can be used to get insight about level of situation awareness. For instance, Kilingaru (2013) used eye gaze analysis to quantify the expertise level of pilot.  Based on the dwell time, duration of fixation of eye on Areas-of-Interest (AOIs), they mapped the cognitive state of the operator into three categories attention focusing, attention blurring, and misplaced attention. Attention focusing is characterized by a continuous dwell on a specific instrument over a period of time with a few glances at other instruments. On the other hand, short dwell time with many saccades between instruments indicates attention blurring. A short dwell on key instruments but with extended dwell outside the instrument panel is an indication of misplaced attention. This work demonstrated that information about pilots’ cognitive states could be obtained by continuous monitoring of the fixation data. Eye tracking has been used in several disciplines to understand situation awareness, expertise level, learning ability and mental state of the human subjects. However, to our knowledge, these opportunities are yet to be explored in the chemical process industries.

 In this study, we used a simulated ethanol plant in a MATLAB environment with an Human Machine Interface (HMI) which closely mimics a typical DCS interface. The experiment is interactive and the participant can use the HMI to get process information and manipulate the control valves. Information about the real-time values of the eleven measured variables, the list of standing alarms, and trend of any one variable (selected by the participant) can be seen in the snapshot of the HMI shown in Figure 1.  The main objective of this work is to understand the cognitive behavior of the operator during disturbances. To accomplish this, the participants (students with control theory background) were given ten separate tasks. During the task, the participants take various actions to assess the situation and eliminate the disturbance by searching for appropriate information and perform corrective actions (opening / closing valves).  These actions taken by the participant using the HMI including mouse movements, mouse clicks and slider movements are recorded during the experiment. TX300 eye tracker is used to record the eye gaze data of the participants. We used this data to calculate dwell time on various AOI to analyze the cognitive behavior of operator abnormal plant situation.

From eye gaze data analysis, we observed that although 91% of the tasks could be successfully controlled by the operators, patterns in the eye gaze behavior indicated that not all successful operations were similar; nearly 12% of the successful operators closely resembled operators who had failed to control the process. They were characterized by relatively lower attention on the primary variables (variables that are directly affected by the disturbance) which indicates a lower understanding of the underlying process dynamics. Dwell time analysis revealed that these participants had relatively lower dwell on the primary trend compared to other participants who had successfully completed the task. Our research also confirmed quantitatively that the use of trend information is important especially in the orientation phase of situation assessment which is in-line with the studies reported by Yin (2012). These insights can be used during operator training to evaluate operators’ expertise in a deeper cognitive level and can also be used to develop an online measure to track the operator’s situation awareness especially during disturbances and abnormal situations.

References

Mannan, Sam, ed. Lees' Loss prevention in the process industries: Hazard identification, assessment and control. Butterworth-Heinemann, 2004.

Gupta, J. P. "The Bhopal gas tragedy: could it have happened in a developed country?." Journal of Loss Prevention in the process Industries 15.1 (2002): 1-4.

Endsley, Mica R. "Design and evaluation for situation awareness enhancement." Proceedings of the Human Factors and Ergonomics Society Annual Meeting. Vol. 32. No. 2. SAGE Publications, 1988.

Hermens, Frouke, Rhona Flin, and Irfan Ahmed. "Eye movements in surgery: A literature review." Journal of Eye Movement Research 6.4 (2013): 1-11.

Kilingaru, K., Tweedale, J.W., Thatcher, S. and Jain, L.C. (2013). "Monitoring pilot “Situation Awareness”." Journal of Intelligent and Fuzzy Systems,24(3), 457-466.

Yin, S., 2012. Proactive monitoring in process contol using predictive trend displays. Ph.D. thesis, Nanyang Technological University, Singapore.


Extended Abstract: File Not Uploaded
See more of this Session: Process Monitoring and Fault Detection II
See more of this Group/Topical: Computing and Systems Technology Division