Qualitative and quantitative analyses of the effects of navigation systems and signal countdown timers on driving behavior

Advances in in-vehicle assisted driving technologies provide new opportunities for improving driving performance and traffic efficiency. Human-machine interfaces (HMIs), navigation systems, and signal countdown timers (SCTs) have changed driver behavior. Although many researchers have attempted to examine the effectiveness of SCT systems, numerous HMI concepts have not been validated. For example, the quantitative impact of SCTs on driving patterns in daily driving tasks remains unclear. Therefore, the objective of this study was to evaluate the effects of an HMI navigation system and an SCT on the driver's visual interaction, performance, and behavioral patterns. Three driving tasks were considered: a baseline (without information prompts), navigation system (with information from a navigation system), and navigation + SCT (with information from a navigation system and an SCT). An experimental study was conducted with 35 participants in a static high-fidelity driving simulator. Data on the driver's visual attention, operating characteristics, and vehicle kinematics were collected. A non-parametric test was used to analyze the human-machine interaction characteristics and driving performance. A Bayesian non-parametric driving model was constructed, and a text clustering algorithm was used to distinguish different driving patterns. The results indicated that providing information from the navigation and SCT systems improved driving performance significantly, whereas providing only information from the navigation system improved driving performance negligibly. Drivers required different information in different traffic light phases, and the most information was needed when the light was green. The proposed driving model could distinguish seven driving patterns during an intersection approach. The quantitative results showed that providing SCT information reduced the probability of acceleration by 21.5% and increased the probability of smooth driving by 25.1%. The findings of the study provide valuable insights for developing user-friendly driver assistance systems.