Babji Srinivasan

In a typical process industry, Human Machine Interfaces (HMIs) are essential for all aspects of communication and interaction between operators and processes, vital for process safety, quality, and efficiency. Good interface design enables operators to accomplish their duties efficiently and effectively with minimal errors. Consequently, it is crucial to design HMIs in a way that facilitates collaboration between automation and operators. Current HMI design practices have a lot of subjectivity and often ignore operator's cognitive behavior. In this work, we propose a quantitative approach that uses cognitive metrics – association and salience metrics – obtained using eye-tracking to improve HMI design for process operators. The association metric helps in identifying information sources which are often used together while salience metric informs about information sources which are extensively used from the HMI. Based on insights from these cognitive metrics, we designed a new HMI. To evaluate the relative efficacy of each HMI, human subject studies were conducted. The results indicate that the new HMI developed with the use of cognitive metrics improved the operators' efficiency. Thus, the proposed approach can help enhance usability of industrial HMIs.

Operator performance is key to ensuring safety in process industries. Therefore, a comprehensive assessment of their performance is critical for smooth and efficient plant operation. Traditional performance assessments are not comprehensive as these ignore cognitive aspects of performance. On the other hand, while eye-tracking-based approaches do provide a cognitive assessment during operating training, their applicability in a real-time real setting is limited. Existing eye-tracking systems come with many constraints affecting users' mobility (restricted movement in all directions) in their environment. In addition, it is beyond the scope of current eye trackers to track multiple users working in an environment, such as in control rooms. Satisfying the above requirements makes these eye-trackers expensive. In this work, we demonstrate the capabilities of an in-house developed cost-effective eye-tracking system to track users' eye movement in an unconstrained environment while giving freedom of head movement. Human subject studies are conducted to compare operators' gaze patterns and the quality of data with that obtained using commercial eye trackers. The results generally agree with data quality obtained using commercial eye trackers. Hence, the performance of the developed eye-tracking system is comparable to existing commercial eye-trackers while overcoming their limitations, such as restricted user movement, single-user tracking, and high cost.