Expanding GOMS-HRA from analog to digital human-machine interfaces

The majority of human reliability analysis (HRA) methods were developed for analog based human-machine interfaces and their applicability to digital HMIs has come in question. Digital HMIs change the nature of the task by presenting information in a windowed interface that requires the user to navigate between displays to access information or controls. GOMS-HRA is a new method developed to analyze human reliability based on the times required to perform tasks in conjunction with the typical human error probability. GOMS-HRA quantifies tasks by decomposing and mapping tasks onto basic task level primitives. It is based off of the goals, objects, methods, and selection rules (GOMS) cognitive model. This paper uses GOMS-HRA to examine operator-in-the-loop study data on nuclear operators interacting with analog and digital HMIs. The results, though inconclusive, suggest digital HMIs may require longer times to complete the same task level primitives compared to analog HMIs. However, this result may be attributed to a deficiency in the GOMS-HRA methodology, since it does not easily account for navigation tasks required for digital HMIs. The GOMS-HRA model was revised to include an additional task level primitive for navigation to allow the methodology to better capture the nuances of interactions for digital HMIs.