729 research outputs found
Sort by
The Relationship between Sensation Seeking and Speed Choice in Road Environments with Different Levels of Risk
No full textThis paper presents the results of a driving simulator study conducted for the UK-funded HumanDrive project, which aims to develop natural, humanlike autonomous vehicle control. As part of that effort, this paper examines whether the established relationship between different sensation seeking (SS) traits and speed choice holds true across a range of driving scenarios, with different levels of contextual risk. Risk was introduced by varying a number of factors, including the environment (rural/urban), and the road edge context (low risk, static risk, potentially dynamic risk). Correlation analysis was performed between sensation seeking and the 95th percentile of vehicle speed for roads with different levels of risk, also considering age and gender. The results indicated that, overall, SS was significantly positively correlated with the 95th percentile of vehicle speed, and particularly for drivers under 40 years. SS was also found to correlate positively with speed choice at all risk levels, however, the effect was more pronounced in road environments that were classified as less risky. These findings have design implications for the development of autonomous vehicle control models
Vehicle Familiarity and Relative Risk of Fatal Crash Involvement
No full textLack of familiarity with a vehicle has been associated with increased crash risk independent of overall driving experience (Perel, 1983). This may pose an underappreciated safety risk in the context of complex and rapidly evolving driver assistance technologies and driver-vehicle interfaces, especially when people drive newly purchased, rented, or borrowed vehicles. The current study estimates the relationship between vehicle ownership and responsibility for crashes using data from 231,056 drivers involved in fatal crashes in the United States in years 2008-2017. A driver was considered responsible for the crash if police indicated that the driver’s pre-crash actions contributed to the occurrence of the crash, and non-responsible otherwise. Driver-, vehicle-, and roadway factors that might also influence crash risk were controlled using logistic regression. Drivers of vehicles registered to another person and drivers of rental vehicles had 1.15 and 1.20 times the odds of responsibility for their crashes, respectively, compared with drivers of their own vehicles. If non-responsible drivers approximate a random sample of all drivers present at the times and places of fatal crashes, these results approximate ratios of responsible involvement in fatal crashes per unit of driving exposure. While ownership is an imperfect proxy for familiarity and may be associated with crash risk by other mechanisms unrelated to familiarity, results are consistent with the hypothesis that drivers of unfamiliar vehicles experience elevated crash risk
Comparison of Virtual Driving Test Performance and On-Road Examination for Licensure Performance: A Replication Study
No full textFor novice drivers, passing the on-road examination (ORE) for licensure marks the transition from supervised to unsupervised driving. However, the first months post-licensure pose the highest lifetime risk of crashing. In partnership with the Ohio Bureau of Motor Vehicles (OBMV), we have developed a virtual driving test (VDT) to enhance new driver skills testing. Through simulation, license applicants were exposed to common serious crash scenarios too dangerous for inclusion in the ORE. In a previous study of an initial sample of 2,143 driver applicants in Ohio, the acceptability, feasibility and construct validity for the VDT was demonstrated: VDT performance variables (simulated traffic collisions and failing to stop at red lights and stop signs) were associated with failing the ORE (all
The Dynamic Merge: Using Traffic Volume Based Signing to Improve Workzone Throughput
No full textRoadwork that results in a lane closure can reduce both the safety and efficiency of a roadway. The dynamic merge is a form of merge control designed to mitigate the potential hazards of lane closures by customizing the merge environment to suit the current level of traffic. When traffic is light, early merge signs encourages drivers to merge into the open lane prior to queue formation. When traffic is heavy, late merge signs encourages drivers to remain in the closed lane for as long as possible. The current study used a driving simulator to assess the independent effects of traffic volume and dynamic merge messaging on merge location and traffic throughput. Merge location was influenced by merge environment, such that drivers in the early merge condition merged earlier than those in the late merge condition regardless of traffic volume. In addition, when traffic was heavy, participants in the late merge condition passed through the work zone more quickly than those in the early merge condition. The results are suggestive of the beneficial effects of the late merge on traffic throughput and the effectiveness of the dynamic merge messaging in influencing merging behavior
Magnetoencephalography during Simulated Driving: A New Paradigm for Driver Assessment
No full textIncreasingly, vehicles are equipped with assistive devices and advanced warning systems to mitigate driver errors, which account for 94% of motor vehicle crashes. However, these technologies require humans to appropriately respond or take over the vehicle. If we want to design effective aids, we need to better understand the neural mechanisms underlying driver error and test how the brain responds to countermeasures. For this, we need sensitive measures of brain activity during driving. This paper present a new paradigm for driver assessment, using magnetoencephalographic (MEG) recording of whole cortex neural oscillatory activity while participants undergo an ecologicallyrelevant simulated driving experience of graded complexity. A pilot experiment set out to demonstrate that expected and motor cortex responses to basic drivingrelated movements (without salient cues) could be recorded, without significant artifact. Following this, a preliminary study of adults (n=5) explored if additional cognitive neural responses to increasing driving task demands can be identified. This paradigm was successfully piloted and preliminary results reveal localized brain regions of expected motor cortex activity, as well as power increases in the frontal lobe. This paradigm can be used to identify not only the neural mechanisms underlying driver errors, but also measure the impact of assistive and alert/warning technologies on these mechanisms in both typical and impaired populations of drivers
Where You Look During Automation Influences Where You Steer After Take-Over
No full textWhen driving a vehicle, gaze direction (where the driver is looking) is tightly coupled with steering actions. For example, previous research has shown that gaze direction directly influences steering behavior. In the context of transitions of control from automated to manual driving, a new question arises: Does gaze direction before a transition influence the manual steering after it? Here we addressed this question in a simplified simulated driving scenario, for maximum experimental control. Participants (N=26) were driven around a constant curvature bend by an automated vehicle, which gradually drifted toward the outside of the bend. An auditory tone cued manual take-over of steering control and participants were required to correct the drift and return to the lane center. Gaze direction was controlled using an onscreen fixation point with a position that varied from trial to trial horizontally and/or vertically. The results showed that steering during manual control was systematically biased by gaze direction during the automated period, but notably in the opposite direction to what might have been expected based on previous research. Whilst further research is needed to understand the causal mechanisms, these findings do suggest that where a driver looks during the seconds preceding a transition to manual control may be critical in determining whether the subsequent steering actions are successful
Hacking Nonverbal Communication between Pedestrians and Vehicles in Virtual Reality
No full textWe use an immersive virtual reality environment to explore the intricate social cues that underlie non-verbal communication involved in a pedestrian’s crossing decision. We “hack” non-verbal communication between pedestrian and vehicle by engineering a set of 15 vehicle trajectories, some of which follow social conventions and some that break them. By subverting social expectations of vehicle behavior we show that pedestrians may use vehicle kinematics to infer social intentions and not merely as the state of a moving object. We investigate human behavior in this virtual world by conducting a study of 22 subjects, with each subject experiencing and responding to each of the trajectories by moving their body, legs, arms, and head in both the physical and the virtual world. Both quantitative and qualitative responses are collected and analyzed, showing that, in fact, social cues can be engineered through vehicle trajectory manipulation. In addition, we demonstrate that immersive virtual worlds which allow the pedestrian to move around freely, provide a powerful way to understand both the mechanisms of human perception and the social signaling involved in pedestrian-vehicle interaction
Autonomous Vehicle Interactions with Other Road Users: Conflicts and Resolutions
No full textAs autonomous vehicles, or AVs, enter the market, other road users will need to interact with them in an effective manner. Currently, in manuallydriven cars, the effectiveness of this interaction is based on the rules of the road that define priorities as well as ad-hoc negotiations to resolve conflicts. To formalize the conflict issue, we introduce the concept of legal zones showing how the road space can be described as graph of these zones. We also introduce the concept of operational regions around a vehicle which must not be infringed upon by others (to avoid safety conflicts). Using these two concepts we show how it is possible to consider new rules for the management of conflict in AV operations. We first briefly describe a new protocol for lane changes and then focus our attention on a protocol for managing conflicts in a pedestrian crossing situation
Consumer Comfort with Vehicle Automation: Changes Over Time
No full textHigher levels of vehicle automation are forecast as a potential mobility solution for many, but understanding consumer comfort and acceptance of selfdriving technologies remains an open question. Results from a series of surveys over three years showed a slight increase in the percentage of people comfortable with full self-driving automation in 2018, following a drop from 2016 to 2017. The recovery in comfort with higher levels of automation was most pronounced among younger adults between ages 25 and 44. However, the percentage of people only comfortable with no automation or features that activate only in certain situations such as in an emergency also increased in the past year, indicating a polarizing trend. Results from the survey also showed that acceptance of self-driving vehicles is conditional on people’s ability to drive as well as having assurance regarding the safety of the technology. Responses also point to a possible misunderstanding among the public regarding the definition and availability of full self-driving technology, indicating a need for improved messaging and consumer education
Dark Personality and Road Crashes: Mediating Role of Driver Vengeance and Violations
No full textAggressive driving and road rage are increasingly leading to Motor Vehicle Collisions (MVC), especially in the developing countries. Considering that malevolent personality characteristics, such as dark triad (narcissism, Machiavellianism, and psychopathy) create a tendency for vengeful and aggressive driving, we examined the power of personality variables in predicting MVC. Specifically, using Contextual Mediated Model (Sümer, 2003), we tested a double mediation model in which driving anger and vengeance mediate the relationships between personality characteristics (Big Five Traits and Dark Triad) and driving violations, and in turn, driving violations mediate the link between driving anger/vengeance and risky driving outcomes (MVC and traffic tickets). Turkish drivers (N = 485, female = %51) completed the measures of personality, aberrant driving behaviors, vengeance, and driving anger. Results of path analyses revealed that whereas narcissism and neuroticism are the critical predictors for aggressive driving Machiavellianism is the strongest predictor of driving vengeance. Moreover, Machiavellianism both directly and indirectly via driving vengeance and violations predicted MVC. Personality variables and mediating variables explained 21% and 26% of the variance in MVC and traffic tickets, respectively, values much higher than those previously reported in the past research. Findings have critical implications for the assessment of aggressive drivers and potential for road rage