Hierarchical Task Analysis for Driving under Divided Attention

Although researchers have made various models of driving behavior, the behavior model under divided attention is not well studied. In this paper, the driver’s behavior differences under divided-attention were studied in a simulated driving environment. A driving scenario was developed to simulate hazards on the highway in dynamic driving conditions. Based on crash and non-crash cases through eye tracking videos from the experiment, Hierarchical task analysis (HTA) was conducted, and decomposed different complex driving behaviors into drivers’ perception, cognition, and decision. Also, their reaction times were compared by using the cognitive-perceptual model in GOMS. Through this study, different driving behaviors and corresponding cognitive factors, which contributed to a slower reaction were identified. The results from this study could be as a valuable input to develop advanced driver assistance systems which could provide smart collision warnings based on the driver's attention.