Towards Establishing Methodology of Naturalistic Driving Simulator Experiment with Multi-agent Traffic Simulation

Jun Tajima (Misaki Design)･Keisuke Suzuki (Kagawa University)･Hironori Suzuki (Toyo University)･Sou Kitajima (JARI)

The realization of realistic virtual traffic environments through multi-agent traffic simulation has enabled the conduct of naturalistic simulator tests, which is antithetical to condition-controlled simulator experiments. Accordingly, this paper discusses the purpose and role of naturalistic simulator tests, their differences from conventional simulator tests, acceptable interventions or operations in experiments, available and effective methods of experimental data analysis.

Analysis of differences in driving behavior between two countries in a naturalistic simulator experiment

Keisuke Suzuki･Naoki Nishiyama (Kagawa University)･Kultida Payappanon･Sunhapos Chantranuwathana･Nuksit Noomwongs (Chulalongkorn University)･Yoshitaka Mimura (Honda R&D)

This paper discusses a methodology for analyzing differences in driving characteristics and risk-taking behavior between the two countries. Using a naturalistic simulator environment proposed by the authors, we analyzed motorcycle risk-taking behavior in Japan and Thailand. Furthermore, we used SEM to visualize the factors that cause behavioral differences and their relationship with human personality.

Information presentation method for preventing speeding of electric kick scooters

Makito Toki (Kagawa University Graduate School)･Keisuke Suzuki (Kagawa University)･Shusaku Toda (Aioi Nissay Dowa Insurance)

To reduce traffic accidents involving electric kick scooters, we analyzed the speed reduction effect of information presentation in a metaverse environment replicating actual urban areas through a naturalistic experiment that replicates normal driving behavior. For this study, we developed a proprietary electric kick scooter simulator and conducted experiments with approximately 20 participants. The results revealed that traveling at 10 km/h significantly reduced collision risk regardless of the risk detection distance. Furthermore, we confirmed that auditory information presentation using beeps had the highest effect in suppressing speeding, while visual information presentation was associated with lower levels of annoyance.

The effect of Camera Monitoring System information presentation on the sense of security of automobile drivers and motorcycle riders

Koki Miyamoto･Keisuke Suzuki (Kagawa University)･Takuya Izumiguchi･Seishi Takagi (Penstone)

We modeled the effects of information presentation via camera monitoring systems on the sense of security of four-wheeled vehicle drivers and two-wheeled vehicle riders using structural equation modeling. Regarding direct effects on the sense of security, we found that for four-wheeled vehicles, the degree of situational awareness significantly influenced this sense, while for two-wheeled vehicles, psychological stability and physical load exerted greater influence than the degree of situational awareness.

Towards Empathetic Interfaces: Effect of Sadness Expression Feedback on Headway Distance Maintenance Characteristics

Toshiaki Kimura (Kyoto Tachibana University)･Mariko shirai (Shinshu University)･Keisuke Suzuki･Tetsuro Butsuen (Kagawa University)･Jun Tajima (Misaki Design)

Emotions are psychological mechanisms that automatically evoke behavior. Therefore, we are conducting research on empathetic interfaces to utilize emotions in promoting safe driving. This paper presents the results of foundational research that experimentally examined the effects of feedback of sad facial expressions to drivers on their headway distance maintenance characteristics. The experiments consist of two parts, preliminary investigation of questionnaire survey and main investigation using driving simulator.

Development of Pedestrian Model of Road Crossing Behavior based on Data obtained from Mixed-Reality Pedestrian Simulator Experiment

Keita Oda (Misaki Design)･Keisuke Suzuki (Kagawa University)･Toshiaki Kimura (Kyoto Tachibana University)･Hironori Suzuki (Toyo University)･Jun Tajima (Setouchi Simulator Inc.)

We observed pedestrian road crossing behavior using Mixed Reality pedestrian simulator, which displays integrated view of real images and computer graphics and allows the pedestrian subjects to walk safely. The simulator tests were conducted under multiple conditions, including the presence or absence of parked vehicles, street trees, and guardrails, as well as varying road width and traffic volumes. The behavior data is analyzed from the perspective of safety attitude. Using the data obtained, discrete choice based pedestrian model is developed to simulate the road crossing behavior of pedestrians, which will be utilized to increase the reality of multi-agent traffic simulation.