Underestimation tendencies of vehicle speed by pedestrians when crossing unmarked roadway

Decision-making of autonomous vehicles in interactions with jaywalkers: A risk-aware deep reinforcement learning approach

Jaywalking, as a hazardous crossing behavior, leaves little time for drivers to anticipate and respond promptly, resulting in high crossing risks. The prevalence of Autonomous Vehicle (AV) technologies has offered new solutions for mitigating jaywalking risks. In this study, we propose a risk-aware deep reinforcement learning (DRL) approach for AVs to make decisions safely and efficiently in jaywalker-vehicle interactions. Notably, a risk prediction module is incorporated into the traditional DRL framework, making the AV agent risk-aware. Considering the complexity of jaywalker-vehicle conflicts, an encoder-decoder model is adopted as the risk prediction module, which comprehensively integrates multi-source data and predicts probabilities of the final conflict severity levels. The risk-aware DRL approach is applied in a simulated environment established in Anylogic, where the motion features of jaywalkers and vehicles are calibrated using real-world survey data. The trained driving policies are evaluated from perspectives of safety and efficiency across three scenarios with escalading levels of jaywalker volume. Regarding safety performance, the Baseline policy performs the worst in "medium jaywalker volume" scenario and "high jaywalker volume" scenario, while our Proposed risk-aware method outperforms the other methods, with the "low TTC ratio" metric stabilizing near 0.08. Moreover, as the scenario gets more complex, the superiority of our Proposed risk-aware policy gets more evident. In terms of efficiency performance, our Proposed risk-aware policy ranks the second best, achieving an "AV delay" metric around 8.1 s in the "medium jaywalker volume" scenario and 8.5 s in the "high jaywalker volume" scenario. In practice, the proposed risk-aware DRL approach can help AV agents perceive potential risks in advance and navigate through potential jaywalking areas safely and efficiently, further enhancing pedestrian safety.

Impacts of brightness contrast, road environment complexity, travel direction and judgement type on speed perception errors among older adult pedestrians' road-crossing decision-making

OBJECTIVES

This study aimed to explore how various factors affect older people's vehicle speed perception to enhance their road safety as pedestrians, focusing on the impact of their cognitive and perceptual abilities on road-crossing decisions.

METHODS

The study evaluated the effects of brightness contrast (high, medium and low), road complexity (high and low) and vehicle travel direction (same and opposite) on speed perception errors in simulated traffic settings. It involved 38 older participants who estimated the speed of a comparison vehicle under two judgement conditions.

RESULTS

Findings showed a consistent underestimation of speed in all conditions. A repeated-measure ANOVA revealed that speed perception errors were significantly higher with low brightness contrast, in simpler road environments, with vehicles travelling in the same direction, and when using absolute judgements.

CONCLUSIONS

These results have practical importance for public safety initiatives, traffic regulation and road design catering to older adults' perceptual needs. They also provide valuable insights for driver training programs for older adults, aimed at enhancing their understanding and management of perceptual biases.

Application of immersive virtual reality for assessing chronic neglect in individuals with stroke: the immersive virtual road-crossing task

BACKGROUND

Neglect can be a long-term consequence of chronic stroke that can impede an individual's ability to perform daily activities, but chronic and discrete forms can be difficult to detect. We developed and evaluated the "immersive virtual road-crossing task" (iVRoad) to identify and quantify discrete neglect symptoms in chronic stroke patients.

METHOD

The iVRoad task requires crossing virtual intersections and placing a letter in a mailbox placed either on the left or right. We tested three groups using the HTC Vive Pro Eye: (1) chronic right hemisphere stroke patients with (N = 20) and (2) without (N = 20) chronic left-sided neglect, and (3) age and gender-matched healthy controls (N = 20). We analyzed temporal parameters, errors, and head rotation to identify group-specific patterns, and applied questionnaires to measure self-assessed pedestrian behavior and usability.

RESULTS

Overall, the task was well-tolerated by all participants with fewer cybersickness-induced symptoms after the VR exposure than before. Reaction time, left-sided errors, and lateral head movements for traffic from left most clearly distinguished between groups. Neglect patients committed more dangerous crossings, but their self-rated pedestrian behavior did not differ from that of stroke patients without neglect. This demonstrates their reduced awareness of the risks in everyday life and highlights the clinical relevance of the task.

CONCLUSIONS

Our findings suggest that a virtual road crossing task, such as iVRoad, has the potential to identify subtle symptoms of neglect by providing virtual scenarios that more closely resemble the demands and challenges of everyday life. iVRoad is an immersive, naturalistic virtual reality task that can measure clinically relevant behavioral variance and identify discrete neglect symptoms.

Prediction of jaywalker-vehicle conflicts based on encoder-decoder framework utilizing multi-source data

Jaywalker-vehicle (J-V) conflicts at mid-blocks without crossing facilities in China are frequent and hazardous. Due to the unexpected and sudden nature of jaywalking activity, it is crucial to develop predictive models for J-V conflicts to offer pre-conflict warnings for road users. This study introduces a novel encoder-decoder framework that utilizes multi-source data to predict J-V conflict severity. We define three encoders to represent three types of input data, (1) J-V interaction encoder (Bi-LSTM), (2) jaywalker motion encoder (Bi-LSTM) and (3) background information encoder (MLP). Subsequently, features extracted by these three encoders are concatenated and transferred to the conflict severity decoder (MLP) to obtain the predicted severity level. We further conduct a case study using the surveyed video data at three mid-blocks without crossing facilities in Nanjing, China. The experimental results indicate that, compared to classical models, our proposed encoder-decoder (Proposed ED) model exhibits the best and stable predictive metrics. Furthermore, the results of the ablation study suggest that the incorporation of background information significantly enhances the four evaluative metrics of the Proposed ED model, with an average improvement of 24.291%. Additionally, the results of transferability analysis suggest that, when the ratio of added samples from the new mid-block reaches 40% to 50%, the predictive metrics of the updated models could stabilize at around 80% to 95%, indicating a notably good performance. Eventually, we derive several practical suggestions from the above findings, in order to help with J-V conflict prediction and jaywalking safety improvement.

Effect of vehicle external acceleration signal light on pedestrian-vehicle interaction

The number of casualties resulting from collisions between pedestrians and motor vehicles continues to rise. A significant factor is the misunderstanding of vehicle behavior intentions by pedestrians. This is especially true with the continuous development of vehicle automation technology, which has reduced direct interaction between drivers and the outside world. Therefore, accurate communication of vehicle behavior intentions is becoming increasingly important. The purpose of this study is to investigate the impact of external vehicle acceleration signal light on the interaction experience between pedestrians and vehicles. The differences between the use and nonuse of acceleration signal light are compared through controlled test track experiments in real scenarios and in videos.The results show that acceleration signal light help pedestrians understand vehicle behavior intentions more quickly and make safer crossing decisions as well as improving their perception of safety when crossing the street and their trust in vehicle behavior.

Smiles and Angry Faces vs. Nods and Head Shakes: Facial Expressions at the Service of Autonomous Vehicles

When deciding whether to cross the street or not, pedestrians take into consideration information provided by both vehicle kinematics and the driver of an approaching vehicle. It will not be long, however, before drivers of autonomous vehicles (AVs) will be unable to communicate their intention to pedestrians, as they will be engaged in activities unrelated to driving. External human–machine interfaces (eHMIs) have been developed to fill the communication gap that will result by offering information to pedestrians about the situational awareness and intention of an AV. Several anthropomorphic eHMI concepts have employed facial expressions to communicate vehicle intention. The aim of the present study was to evaluate the efficiency of emotional (smile; angry expression) and conversational (nod; head shake) facial expressions in communicating vehicle intention (yielding; non-yielding). Participants completed a crossing intention task where they were tasked with deciding appropriately whether to cross the street or not. Emotional expressions communicated vehicle intention more efficiently than conversational expressions, as evidenced by the lower latency in the emotional expression condition compared to the conversational expression condition. The implications of our findings for the development of anthropomorphic eHMIs that employ facial expressions to communicate vehicle intention are discussed.

Lane-based Distance-Velocity model for evaluating pedestrian-vehicle interaction at non-signalized locations

Pedestrian vehicle conflicts at non-signalized crosswalks are a world-wide safety concern. Although the "pedestrian priority" policy is applied in some regions to improve pedestrian safety, its effect needs further investigation. This study proposes the Lane-based Distance-Velocity model (LDV) to investigate pedestrian-vehicle interaction at non-signalized crosswalks. Compared with the DV model, the LDV model considers the lateral distance between vehicles and pedestrians. Therefore, the LDV model extends the application of the DV model by allowing it to be applied not only on one-lane streets to multi-lane streets. The conflict severities of pedestrian-vehicle interaction in the LDV model are classified into four categories: safe-passage, mild-interaction, potential-conflict and potential-collision. Based on that, pedestrian crossing decisions are graded as safe-crossing, risky-crossing, and dangerous-crossing. The experimental data are collected at a non-signalized crosswalk through drone footage collected in Xi'an City (China) with a Machine Vision Intelligent Algorithm. The model is tested through a case study to evaluate pedestrian crossing safety when interacting with private cars and taxis. Results from the case study suggest that the proposed model works well in the pedestrian-vehicle interaction analysis. Firstly, 87.9% of drivers are willing to provide right-of-way to pedestrians when they have enough time to react and yield. Then, both the DV model and LDV model have reached consistent conclusions: the deliberate violation rate (DVR) of taxi drivers is 22.64%, which is double that of private car drivers. Last, taxis commit a higher percentage of pedestrians' dangerous or risky crossing situations than private cars. Relevant government departments can utilize the results of this study to manage urban traffic better and improve pedestrian safety.

Ghost on the Windshield: Employing a Virtual Human Character to Communicate Pedestrian Acknowledgement and Vehicle Intention

Pedestrians base their street-crossing decisions on vehicle-centric as well as driver-centric cues. In the future, however, drivers of autonomous vehicles will be preoccupied with non-driving related activities and will thus be unable to provide pedestrians with relevant communicative cues. External human–machine interfaces (eHMIs) hold promise for filling the expected communication gap by providing information about a vehicle’s situational awareness and intention. In this paper, we present an eHMI concept that employs a virtual human character (VHC) to communicate pedestrian acknowledgement and vehicle intention (non-yielding; cruising; yielding). Pedestrian acknowledgement is communicated via gaze direction while vehicle intention is communicated via facial expression. The effectiveness of the proposed anthropomorphic eHMI concept was evaluated in the context of a monitor-based laboratory experiment where the participants performed a crossing intention task (self-paced, two-alternative forced choice) and their accuracy in making appropriate street-crossing decisions was measured. In each trial, they were first presented with a 3D animated sequence of a VHC (male; female) that either looked directly at them or clearly to their right while producing either an emotional (smile; angry expression; surprised expression), a conversational (nod; head shake), or a neutral (neutral expression; cheek puff) facial expression. Then, the participants were asked to imagine they were pedestrians intending to cross a one-way street at a random uncontrolled location when they saw an autonomous vehicle equipped with the eHMI approaching from the right and indicate via mouse click whether they would cross the street in front of the oncoming vehicle or not. An implementation of the proposed concept where non-yielding intention is communicated via the VHC producing either an angry expression, a surprised expression, or a head shake; cruising intention is communicated via the VHC puffing its cheeks; and yielding intention is communicated via the VHC nodding, was shown to be highly effective in ensuring the safety of a single pedestrian or even two co-located pedestrians without compromising traffic flow in either case. The implications for the development of intuitive, culture-transcending eHMIs that can support multiple pedestrians in parallel are discussed.

An integrated text mining, literature review, and meta-analysis approach to investigate pedestrian violation behaviours

The goal of this study is to provide an overview of previous research that investigated pedestrian violation behaviour, with a focus on identifying the contributing factors of such behaviour, its impact on pedestrian safety, the mitigation strategies, the limitations of current studies, and the future research directions. To that end, the Latent Dirichlet Allocation (LDA) text mining method was applied to extract a comprehensive list of studies that were conducted during the past 21 years related to pedestrian violation behaviours. Using the extracted studies, a multi-sectional literature review was developed to provide a comprehensive understanding of the different aspects related to pedestrian violations. Afterward, a meta-analysis was undertaken, using the studies that reported quantitative results, in order to obtain the average impact of the different contributing factors on the frequency of pedestrian violations. The study found that pedestrian violations are one of the hazardous behaviours that contribute to both the frequency and severity of pedestrian-vehicle collisions. According to the literature, the waiting time at the curbside, traffic volume, walking speed, pedestrian distraction, the presence of bus stops and schools, and the presence of on-street parking are among the key factors that increase the likelihood of pedestrian violations. The study has also reviewed a wide range of strategies that can be used to mitigate violations and reduce the safety consequences of such behaviour, including simple engineering-based countermeasures, enforcement, solutions that rely on advanced in-vehicle technologies, and infrastructure connectivity features, educational programs, and public campaigns.

Investigating yielding behavior of heterogeneous vehicles at a semi-controlled crosswalk

It is well known that pedestrians are vulnerable road users. Their risk of being injured or killed in road traffic crashes is even higher as vehicle drivers often violate traffic rules and do not slow down or yield in front of crosswalks. In order to reduce this risk, many countries have issued strict regulations requiring vehicles to yield to pedestrians in front of crosswalks. While extensive literature exists on the interaction between vehicles and pedestrians, the consideration of heterogeneity in the behavior of vehicles is vastly overlooked. Accordingly, this study analyzes the yielding behavior of three types of vehicles under the "pedestrian priority" policy by processing drone footage collected in Xi'an City (China) with a Machine Vision Intelligent Algorithm. Moreover, this study proposes four additional indicators to the widely used yielding rate and yielding delay with the aim of evaluating yielding behavior of three types of vehicles. The results show that buses have the best yielding behavior from the perspective of yielding rate, yielding delay, waiting time, yielding angle and waiting site. Buses perform well in observing pedestrian dynamics near crosswalk, and perform exceptionally well in considering the "blind area" of vision. The location of the waiting site in front of the stop line and the length of the waiting time contribute to the safe crossing of pedestrians. In contrast, private cars perform badly in yielding to pedestrians. However, serious polarization can be observed across private cars, as the performance varies across the board. The relaxation of the homogenization assumption of the behavior of vehicles in pedestrian-vehicle interaction, alongside the improvements in the analysis via Machine Vision Intelligent Algorithm of videos acquired via drone, shows the possibility of having a deeper understanding of the yielding behavior of vehicles at crosswalk. The extension of the use of artificial intelligence methods to analyze drone footage has immense potential in understanding road user behavior and hence providing knowledge for crash prevention.

Equestrian Road Safety in the United Kingdom: Factors Associated with Collisions and Horse Fatalities

Over 60% of UK horse riders report having experienced a road-related near-miss or accident. The aim of this study was to describe horse-related road incidents (n = 4107) reported to the British Horse Society (2010-2020) and to identify factors associated with higher odds of collisions with another vehicle and horse fatalities using multivariable logistic regression modelling. Drivers passed the horse too closely in 84.2% of incidents while road rage and speeding were reported in 40.3% and 40.1% of incidents, respectively. Close passing distance alone (odds ratio [OR] 18.3, 95% confidence interval [CI] 6.5, 51.6) or in combination with speeding (OR 4.4, CI 1.7, 11.7) was associated with higher collision odds compared to speeding alone. Speeding was, however, associated with higher horse fatality odds (OR 2.3, CI 1.2, 4.6). Wearing high visibility clothing reduced odds of collision (OR 0.2, CI 0.1, 0.4). A fatal injury to a horse was almost 12 times as likely to result in severe to fatal rider/handler injury. Loose horses contribute significantly to road-related horse fatalities. Driver behaviour of how to pass horses safely on UK roads needs further improvement and will help reduce the risk of collisions and horse and human fatalities.