Promotion of Proactive Behavioral Change to Prevent Human Error
-Part 1 Longitudinal relationships between perceptual–cognitive functions and strategic driving behavior in older drivers-

Toshihisa Sato (AIST)･Hiroshi Yoshitake (Institute of Science Tokyo)･Yoshiko Goda･Masashi Toda (Kumamoto University)･Motoki Shino (Institute of Science Tokyo)

This paper describes the relationship between perceptual and cognitive functions and strategic driving behavior at two time points, one year and four years later. The study involved 106 older drivers (57 of whom were female). Structural equation modeling contributed to identifying the strategic driving behaviors where the impact of changes in cognitive function over time became evident, as well as those where the cognitive factors influencing them changed.

Promotion of Proactive Behavioral Change to Prevent Human Error
-Part 2 Analysis of Behavioral Change Processes in Older Drivers Through Collaborative Reflection-

Yoshiko Goda･Masashi Toda (Kumamoto University)･Maki Arame･Junko Handa (The Polytechnic University of Japan)･Hiroshi Yoshitake (Institute of Science Tokyo)･Toshihisa Sato (AIST)･Motoki Shino (Institute of Science Tokyo)

This study analyzed the process of safe behavioral change by combining driving data from AI-equipped training vehicles with reflective dialogue among nine older drivers. Participants collaboratively reflected on their driving characteristics and hazardous situations through group discussions. The results indicated a reduction in major driving errors, along with increased awareness of personal driving tendencies, perspective changes through interaction with others, and metacognitive reflections related to hazard prediction. This paper qualitatively examines how collaborative reflection contributes to proactive behavioral change in older drivers.

Promotion of Proactive Behavioral Change to Prevent Human Error
-Part 3 Effects of Presenting Benefits of Behavioral Change in Driving Behavior Intervention for Older Drivers-

Hiroshi Yoshitake (Institute of Science Tokyo)･Jongseong Gwak (Takushoku University)･Motoki Shino (Institute of Science Tokyo)

To promote voluntary changes in driving behavior among older drivers, a previous study proposed first-person perspective training with feedback. However, individual differences in intervention effects suggested that understanding the rationale and benefits of behavioral change may influence outcomes. In this study, we presented information on the rationale and benefits of changing driving behavior during the intervention and examined how the content of this information affected post-intervention changes in driving behavior.

Promotion of Proactive Behavioral Change to Prevent Human Error
-Part 4 Estimation of Driver's Head Pose, Gaze, and Driving Posture and Visualization of Gaze Targets Using In-Vehicle Cameras for Driving Behavior Analysis-

Ryota Ishihara･Akito Yamasaki (Meijo University)･Hiroshi Yoshitake･Motoki Shino (Institute of Science Tokyo)

Quantitative analysis of driving behavior is essential for developing assistance strategies that promote drivers' autonomous behavioral change. In this study, we propose a system that estimates a driver's head pose, gaze direction, and driving posture from in-vehicle camera images in a non-contact manner, and visualizes gaze targets by projecting the gaze direction onto the forward-view image. The effectiveness of the proposed method is demonstrated through experiments in an actual vehicle environment.

Promotion of Proactive Behavioral Change to Prevent Human Error
-Part 5 Development of a Fundamental Technology for Integrating Attention Targets Extracted via LLM-based Traffic Context Understanding with Driver Gaze Targets-

Soma Yamasaki･Akito Yamasaki (Meijo University)･Hiroshi Yoshitake･Motoki Shino (Institute of Science Tokyo)

Conventional driving behavior analysis can evaluate "what the driver looked at," but it has been difficult to address the requirements imposed by the traffic environment, namely "what the driver should have looked at." In this study, we propose a method that extracts and structures "attention targets" from road environment images using a large language model, and integrates them with the driver's "gaze targets" to quantitatively evaluate the appropriateness of confirmation behavior.

Development of Driver Education System for Calibrating Older Drivers' Self-Assessment

Motoki Shino･Shuto Inouchi･Hiroshi Yoshitake (Institute of Science Tokyo)･Ryota Fujita･Hiroto Kato (Mitsubishi Precision)

We developed a driver education system designed to calibrate drivers’ self-assessment of their driving without the involvement of an instructor utilizing a driving simulator. The system was applied to older drivers, and the results suggested that it can increase safety awareness and anxiety about traffic accidents, while improving driving behavior at intersections. The results also revealed that greater correction of overestimation through driver education was associated with greater improvement in driving behavior.