論文

Human behavior and interaction in road traffic is highly complex, with many open scientific questions of high applied importance, not least in relation to recent development efforts toward automated vehicles. In parallel, recent decades have seen major advances in cognitive neuroscience models of human decision-making, but these models have mainly been applied to simplified laboratory tasks. Here, we demonstrate how variable-drift extensions of drift diffusion (or evidence accumulation) models of decision-making can be adapted to the mundane yet non-trivial scenario of a pedestrian deciding if and when to cross a road with oncoming vehicle traffic. Our variable-drift diffusion models provide a mechanistic account of pedestrian road-crossing decisions, and how these are impacted by a variety of sensory cues: time and distance gaps in oncoming vehicle traffic, vehicle deceleration implicitly signaling intent to yield, as well as explicit communication of such yielding intentions. We conclude that variable-drift diffusion models not only hold great promise as mechanistic models of complex real-world decisions, but that they can also serve as applied tools for improving road traffic safety and efficiency.

Automated vehicles (AVs) are expected to be used as mobility services for improving the quality of life in aging rural areas. In this study, we investigated problematic cases of interaction between pedestrians and AVs in Japanese rural areas and observed cases that occurred on narrow roads. To explore a better communication method when pedestrians interact with vehicles, a virtual reality experiment examined the impacts of two communication methods (external human-machine interface and vehicle movement) regarding two types of vehicles (golf cart and bus). The results showed that pedestrians could decide on crossing the road quickly when the AV attempted to reduce speed early and stopped maintaining an appropriate distance from pedestrians. The stopping position is an important contributor to pedestrian decision-making in crossing and attitude toward AVs. The findings of this study have implications for the design of future automated service vehicles in rural areas.

Due to the technical advances of automated vehicles (AVs), new uncertainties for human road users arise. To overcome these uncertainties, driving strategies of AVs might be aligned to human interaction styles. In vehicle-vehicle interactions, driving behavior is informed by remaining time gaps between vehicles. This video-based experiment investigated the influence of gap sizes and the measurement method on driving decisions. N=32 participants experienced a highly automated drive in which their AV approached narrow passages. The time to arrival (TTA) of the oncoming traffic was varied. Participants had to decide to drive first or second, indicate their decision certainty, and the situation's criticality. The videos were presented in ascending, descending, and random order. Moreover, participants adjusted the TTA at which they would drive first and second. The results indicated a higher probability of driving first and lower criticality with increasing TTA. Decision certainty was lowest around the 50% threshold, while longer and shorter TTAs resulted in higher certainty. Results differed between the methods. The findings provide guidance for the design of automated systems to mimic human driving behavior.

For an appropriate communication design between an automated vehicle and road users, we need to understand unsafe and inefficient communications in public road environments. This study investigated unsafe and inefficient communications between automated buses and road users and analyzed the underlying factors. We collected video data from a camera attached to a bus for 233 days in a field operation test (FOT) at seven locations in Japan and observed 22,199 communications between the automated bus and road users. Consequently, we observed several types of unsafe and inefficient communications. We found that specific automated bus characteristics, such as absence of driver's action, unfamiliar appearance, and fixed trajectory, caused these unsafe and inefficient communications in crossing and overtaking scenarios. Our study indicates the necessity of some improvements for implicit/explicit cues from an automated bus, along with the education of residents and visitors.

This study explored communication designs for driverless automated service vehicles (ASVs) and pedestrians' trust and perceived safety when pedestrians cross narrow roads in residential areas. Forty participants carried out VR experiments that simulated interactions between implicit (vehicle behavior) and explicit communications (text-based external human-machine interfaces (eHMI)) of two ASVs types (bus and golf cart) in residential areas. Results indicated that pedestrians were more likely to trust the bus-type ASV more when it stopped further away compared to decelerating relatively late. Pedestrians' perception of safety varied depending on the type of ASV, with the golf cart type being perceived as safer than the bus type. Further, pedestrians exhibited greater trust when the ASV did not display any messages. However, text messages indicating the ASV's status, such as”In Automated Driving,” improved their perception of safety. This highlights the need for effective communication methods to enhance road user attitudes towards ASVs.

This study aimed to investigate how human, vehicle, and environment (HVE)-related factors and their interactions contribute to fall accidents related to electric scooters (e-scooters). Falls are the most common type of e-scooter accidents, and developing a thorough understanding of the factors that contribute to these accidents is critical for effective accident prevention. Unlike collisions, falls frequently result from the complex interaction among the rider, the vehicle, and the environment. To this end, this study conducted a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and uses the Haddon Matrix framework to identify and classify factors related to e-scooter fall accidents from HVE perspectives, spanning the pre-fall and fall phases. The findings suggest that e-scooter fall accidents are multifactorial, resulting from the interaction of HVE-related factors across accident phases rather than from a single cause. Human-related factors, vehicle attributes, and environmental conditions were all found to contribute to fall risk, with notable interactions identified across all three dimensions. This study contributes to a better understanding of the mechanisms underlying e-scooter fall accidents by systematically identifying these factors and examining their interactions, highlighting the need for further investigation into HVE interactions across diverse accident contexts.

Communication between pedestrians and automated vehicles is playing a key role in enhancing the safety of future traffic environment. The current study attempted to suggest new insights into designing external human-machine interfaces (eHMIs) in automated vehicles for traffic safety as examines negative effects of the eHMI on pedestrian crossing behaviour in a situation where an automated vehicle yields to pedestrian. Virtual Reality systems simulated three experimental conditions: baseline (no eHMI), showing “After you” and “I’ll stop” via eHMI on an automated vehicle in residential areas. The experiment using human participants resulted that conveying information via eHMI led pedestrians to do less careful exploratory behaviour toward other traffic. The result also showed the greater number of traffic collisions when the eHMI showed information compared to non-eHMI condition. The findings of this study are also being used to help how to design the eHMI on automated vehicles in shared spaces.

The introduction of driving automation as a mobility service has been expected to shape new types of road communication between vulnerable road users and automated vehicles. Given that road users' lack of knowledge about automated vehicles in depopulated rural areas, it is important to design conversative strategies leading effective pedestrian-automated vehicle communication. The current study investigates impacts of communication methods assisting pedestrian avoidance from an approaching automated driving golf cart from behind. Three communication methods (Baseline, text-based external human-machine interface, road marking) and two environments (narrow road, parking area) were simulated in virtual reality experiments. Results found that pedestrians found it useful and effective when the automated cart provided a text message via the external human-machine interface, with giving way to the automated cart. As the blue marking on the road indicated the driving route of the automated cart, the road marking also had a potential for leading pedestrians to move from the road. Empirical findings provide practical recommendations for the design of communication strategies leading pedestrian-automated vehicle interaction in depopulated areas.

In this study, we examine the differences in pedestrian behavior at crosswalks between communicating with conventional vehicles (CVs) and automated vehicles (AVs). To analyze pedestrian behavior statistically, we record the pedestrian’s position (x- and y-coordinates) every 0.5 s and perform a hot spot analysis. A Toyota Prius (ZVW30) is used as the CV and AV, and the vehicle behavior is controlled using the Wizard of Oz method. An experiment is conducted on a public road in Odaiba, Tokyo, Japan, where 38 participants are recruited for each experiment involving a CV and an AV. The participants cross the road after communicating with the CV or AV. The results show that the pedestrians can cross earlier when communicating with the CV as compared with the AV. The hot spot analysis shows that pedestrians who communicate with the CV decide to cross the road before the CV stops; however, pedestrians who communicate with the AVs decide to cross the road after the AV stops. It is discovered that perceived safety does not significantly affect pedestrian behavior; therefore, earlier perceived safety by drivers’ communication and external human–machine interface is more important than higher perceived safety for achieving efficient communication.

One of the challenges in introducing automated service vehicles (ASVs) to parking areas is guiding pedestrians to make quick and appropriate crossing decisions based on an understanding of what the ASV typically does. This study aims to find effective communication strategies for ASVs for pedestrian crossings in Japanese parking areas. Virtual Reality experiments were conducted with 40 participants facilitating vehicle behavior (Normal Deceleration, Early Deceleration, Early Stop) and text-based external human-machine interfaces (eHMIs; No eHMI, “After you,” “I’ll stop,” “In Automated Driving”) as tools of communication with two types of ASVs (bus and golf cart) in parking areas. The results indicated that vehicle behavior significantly influenced pedestrian crossing decisions, and pedestrians began to cross when the ASV decelerated early and stopped not too far from them (Early Deceleration). Self-reported measure analyses revealed that text messages containing the intent of the ASV's actions or instruction (After you, I’ll stop) with Early Deceleration would probably lead to pedestrians’ favorable attitudes toward automated buses. These findings contribute to communication designs based on implicit communication for pedestrian crossing decision-making in parking areas with a clear interpretation of ASV’s intentions.

An external human-machine interface (eHMI) is a novel tool that facilitates efficient interactions between pedestrians and automated vehicles (AVs). However, as pedestrians become accustomed to communicating with AV through eHMI, there is a risk of exhibiting less cautious behavior, which can increase the likelihood of accidents. This study aimed to examine the effects of methods to mitigate the side effects of communication with AV through an eHMI. A total of 132 participants participated in a Virtual Reality experiment to investigate the existence of negative effects and to explore mitigation methods. We designed five experimental conditions to manipulate the tone of text messages (Allocentric: “After you,” Egocentric: “I’ll stop”) and test the effect of mitigation methods, such as hiding the text messages on the eHMI when the AV is stopped (No eHMI, eHMI-A, eHMI-E, HeHMI-A, HeHMI-E). The results revealed that pedestrians were more likely to exhibit less careful behavior in the eHMI condition than in the No eHMI condition. The disappearance of text messages on the eHMI prompted pedestrian attention allocation toward traffic situations, leading to a reduction in accident risk. In particular, the mitigation method was most effective for safe crossing when pedestrians were continuously exposed to the eHMI, presenting an egocentric message. Our findings contribute to the design of text-based eHMIs for pedestrian decision-making when crossing and enhancing traffic safety.

In the near future, pedestrians will face highly automated vehicles on the roads. Highly automated vehicles (HAVs) should have safety-enhancing communication tools to guarantee traffic safety, e.g., vehicle kinematics and external human–machine interfaces (eHMIs). Pedestrians, as highly vulnerable road users, depend on communication with HAVs. Miscommunication between pedestrians and HAVs could quickly result in accidents, and this, in turn, could cause severe impairments for pedestrians. Light-band eHMIs have the potential to enhance traffic safety. However, eHMIs have been less explored in Japan so far. As a first-time approach, this experimental online study shed light on the effect of a light-band eHMI on Japanese pedestrians (N=99). In short video sequences, the participants interacted with two differently sized HAVs equipped with light-band eHMI. We investigated the effect of vehicle size (small vs. large), eHMI status (no eHMI vs. static eHMI vs. dynamic eHMI), and vehicle kinematics (yielding vs. non-yielding) on pedestrians’ willingness to cross, trust, and perceived safety. To investigate possible side effects of eHMIs, we also included experimental conditions in which the eHMI mismatched the vehicle's kinematics. Results revealed that Japanese were more willing to cross the street and indicated higher trust- and safety ratings when they received information about the vehicle's intention and automation status (dynamic eHMI) compared to when they received no information (no eHMI) or only about the vehicle automation status (static eHMI). Surprisingly, Japanese participants tended to rely on the eHMI when there was mismatching information between eHMI and vehicle kinematics. Overall, we concluded that light-band eHMIs could contribute to a safe future interaction between pedestrians and HAVs in Japan under the requirement that the eHMI is in accordance with vehicle kinematics.

Crossing strategy refers to pedestrian objectives in choosing crossing patterns and their pace. To the best of our knowledge, there remains a deficiency in literature regarding pedestrian crossing strategies when interacting with autonomous vehicles (AVs). This study investigated the effect of various physical road infrastructures on pedestrian crossing strategies while interacting with multiple AVs. The effects of three infrastructures in a typical local street on the crossing strategies of pedestrians were identified using a structural equation model based on a stimulus–organism–response (SOR) framework. The stimulus included sidewalk, crosswalk, and/or legal on-street parking designed virtually in a four-lane local intersection, where frequent daily pedestrian–AV interactions occurred in a mixed neighborhood. The organism dimension of pedestrians was measured in terms of situation awareness (SA) and perceived risk (PR) when interacting with multiple AVs. Pedestrian crossing strategies, including their intended crossing patterns and speeds, were identified in the response dimension. Based on experimental data from 82 university students, the findings revealed that sidewalk and legal on-street parking significantly affected SA with coefficients of 0.155 and − 0.079, respectively, whereas the crosswalk had a remarkable association with PR by a coefficient of −0.098. In addition, there was a trade-off relationship between pedestrian patterns and speed (coefficient of −0.197) when interacting with multiple AVs. Our research on pedestrians interacting with multiple AVs provides novel insights to enhance the understanding of pedestrian crossing strategies and establish a comparison with the realities of crowded local roadways. These insights expand our knowledge of actual pedestrian crossing behaviors in the AV environment and support the production of safer street design guidelines.

© Springer Nature Switzerland AG 2019. Landmarks and clues outside the car that elderly drivers can easily recognize from inside the car and on a car navigation system during route navigation were extracted. In a laboratory survey, we measured the degree of awareness of logos by type of business. We then conducted a driving simulator experiment to verify the response rate, reaction distance, and line of sight direction according to the business type and store, vehicle location, shop location, and presence/absence of signboards that had good results in the laboratory survey set as variables. Results show that a vehicle navigation system should ideally use clue information outside the car that drivers can view without largely averting their eyes from the front; use clue information that is presented at a large visual angle; and use clue information that is conspicuous in color and shape and known by the elderly driver.

The interaction between a pedestrian and an automated vehicle equipped with an external human machine interface at an unsignalized crosswalk is discussed. The external human machine interface has the potential to provide an effective communication cue from which the pedestrian can judge whether the automated vehicle is yielding to them.