729 research outputs found
Sort by
Effects of Behavior-Based Driver Feedback Systems on Commercial Long Haul Operator Safety
No full textThere are large economic and societal costs to commercial motor vehicle crashes. A majority of crashes are precipitated due to driver-related factors. Behavior-based systems that influence drivers with feedback from safety managers can help reduce driver-related risk factors. These systems harness the experience and knowledge of managers along with advanced driver telematics that monitor and record driver behaviors to positively influence driver safety. Safety solutions that focus on modifying driver behaviors thus hold promise for improving the safety record of commercial trucking. In this study, one such feedback system was examined by analyzing data from a commercial trucking fleet, treating the system deployment as a natural experiment. This made it possible, without experimental intervention, to compare drivers before and after system introduction, and to compare drivers that were subject to this system with those that drove with no supervisor feedback. Adverse event data were obtained for drivers in the fleet and weekly event rates were calculated taking into account driving exposure (in miles). Results show that drivers improved after receiving safety feedback and significantly more so than drivers that did not receive feedback
Are Rear-End Crashes Caused Mainly by an Interaction Between Glance Duration and Closure Rate?
No full textVictor and colleagues recently analyzed rear-end crash and near-crash data from the Strategic Highway Research Program Phase 2 naturalistic driving study. They measured the last off-path glance duration just before a crash, and the closure rate (the change in looming rate during the last glance). They concluded that the predominant cause of rear-end crashes was a “mismatch”– a short glance with a high closure rate or a long glance with a low closure rate. The current study independently tested this “mismatch” hypothesis using two epidemiological methods to estimate odds ratios for combined crashes and near-crashes (events). First, the glance and closure rates were stratified, and compared to a “just following,” no-crash baseline (short glances 0.056 s-2, glances increased event risk, proportional to glance duration. However, a major data limitation that potentially upwardly biased the interaction OR estimate is an inherent dependency of closure rate on glance duration, simply because of the way closure rate was defined. The mismatch hypothesis for rear-end events must be tested with other rear-end event datasets not subject to this limitation before being considered fully validated
Dynamic Workload Fluctuations in Driver/Non-Driver Conversational DYADS
No full textWe developed a new method for simultaneously assessing the workload of a driver and a non-driver engaged in natural conversation either in the vehicle or over a cell phone. For both the driver and non-driver, talking was found to be more demanding than listening and the pattern was identical for both passenger conversations and cell phone conversations. Operating the vehicle increased the workload for the driver over and above the conversation task. The effects of driving (or not) and talking (or not) were found to be additive. The data reveal a pattern of dynamic fluctuation in workload in driver/non-driver conversational dyads
It’s Out of Our Hands Now! Effects of Non-Driving Related Tasks During Highly Automated Driving on Drivers’ Fatigue
No full textWith introduction of conditional automation in vehicles the driver can engage in non-driving related tasks (NDRTs) and only has to intervene in case of take-over requests (TOR). Therefore, active fatigue, which is the most frequent form of fatigue in manual driving, is assumed to be replaced by passive fatigue, intensified through monotony and monitoring tasks in conditional automated driving (CAD), SAE Level 3. To investigate effects of NDRTs on drivers’ fatigue and take over capability a driving simulator study was conducted. In total, 56 participants experienced two rides on a highway with CAD. During the two rides, participants had to fulfill both a monotonous monitoring task and an activating task. As in CAD the system is executing longitudinal and lateral control, drowsiness detection referring to driving performance becomes inoperative. Noninvasive methods for drowsiness detection that are not related to driving performance have to be investigated. Therefore, fatigue was measured with percentage of eye-lid closure (PERCLOS), blink related eye-tracking parameters, and the self-report Karolinska Sleepiness Scale (KSS). Results suggest that fatigue can be caused through a monitoring task in highly automated driving. PERCLOS could be confirmed as a valid parameter for detecting fatigue in CAD. Further, passive task related fatigue caused by a 25 min monotonous monitoring task does not affect the drivers’ take over capability negatively
Effectiveness of Visual Collision Warning Alerts on Young Drivers’ Latent Hazard Anticipation
No full textForward roadway collision warning systems can reduce rear-end collisions, among other unsafe behaviors. Previous studies have shown that young drivers fail to scan adequately for latent hazards. The current driving simulator study investigates the effect of visual collision warning messages on drivers’ hazard anticipation ability, when presented either 2s, 3s or 4s in advance of a potential threat. This experiment examined the latent hazard anticipation behavior of forty-eight young drivers aged 18-25 across eight unique scenarios both, in the presence, and absence of visual collision warning alerts. The analysis of glance data captured using an eye tracker, show that visual warning messages significantly increased the proportion of latent hazards anticipated regardless of hazard type (pedestrian or vehicle). The 2s warning duration was found to statistically have the same effect on hazard anticipation compared to the 3s and 4s warning thresholds. The study has potential implications for the effective design of forward collision warning systems
A Preliminary Examination of Age-Related Differences in Perceived Complexity at Fatal-Crash Intersections
No full textYounger and older drivers are both overrepresented in fatal crashes that occur at intersections, however, after adjusting for other significant factors (i.e., being at fault type of road, weather, lighting) the increased risk cannot be fully accounted for older drivers, nor does frailty. Thus, increased risk for older drivers could be due to their agerelated cognitive declines and possible differences in perceptions of intersections. The current study examines whether older drivers’ perceived complexity of intersections differed quantitatively and qualitatively from younger drivers’ perceived complexity of the same intersections. Coordinates of a random sample of intersections where at least one fatality occurred over a three-year period from the US Fatality Analysis Reporting System (FARS) were identified and Google Earth was used to extract still images of each intersection. The complexity of these intersection images were rated by a sample (N =38) of younger (age 18-35) and older drivers (age 65+). Inter-rater reliability for each group was calculated. In addition, individual intersection images with the largest and smallest age differences were qualitatively examined. Results suggest that older drivers view the complexity of intersections differently than younger drivers. Overall, older drivers were less reliable and scored nominally higher on average in their complexity ratings than younger drivers. Moreover, older drivers tended to rate rural or residential intersections as being more complex than younger drivers; whereas younger drivers tended to rate urban intersections as being more complex. Future work should account for these age differences in perceived intersection complexity
Comparing Methods of Detecting Mind Wandering While Driving
No full textDriver distraction is a persistent threat to traffic safety. External distraction has been examined extensively, but few studies have focused on internal distraction such as mind wandering. Equivocal results from the few existing studies are likely due, at least in part, to different experimental methods. Mind wandering is commonly assessed using either a self-caught or probe-caught method. The current investigation sought to better understand the effects of mind wandering on driving performance using the self-caught method and the probecaught method. In the Self-Caught Experiment, lateral control measures such as, lateral position variability and steering reversal rate were greater when drivers reported on-task thoughts versus mind wandering. In the Probe-Caught Experiment, these results were not replicated using the traditional probe-caught analysis. Instead, when analyzing the results of the Probe-Caught Experiment in a similar manner as the Self-Caught Experiment, the results were replicated. These results highlight methodological concerns in detecting mind wandering while driving. Additional research is needed to determine which method should be employed in future studies
Toward an Antiphony Framework for Dividing Tasks into Subtasks
No full textTask analysis is a staple of ergonomics, neuroergonomics, human factors, and experimental psychology inquiry, and often benefits from granularity beyond the task level to the subtask level. The concept and challenge of identifying the subcomponents of tasks are neither new, nor solved. Practitioners routinely identify individually internally consistent and yet conflicting subdivisions. The challenge of producing reliable, valid subtask data across efforts recommends a unified framework for identifying consistent subtask divisions within tasks. A framework is here forwarded, based upon universal “antiphony” turn-taking behavior in human-human interaction, but adapted to address the highly scripted vocabulary of human-machine interaction. Practical application to a real-world vehicle interface is demonstrated, an example discussed in the light of research design, applied use, and future improvement
Training to Improve Collision Detection in Older Adults
No full textPrevious studies have indicated a decline in the ability of older adults to detect impending collisions. In addition, previous research has demonstrated that collision detection performance of college-aged participants can be improved with perceptual learning. The present study examined whether perceptual learning can improve performance of older participants on a collision detection task (N=12). The experiment was conducted over seven days with each day consisting of a 1-hr session. Thresholds for three observer speeds were measured prior to training using a two-alternative forced choice procedure during which participants indicated whether an approaching object would result in a collision or noncollision event. Participants were then trained near threshold at one of these speeds for 5 days. After training participants’ thresholds were measured again. Results indicate a significant reduction in the time needed to detect a collision for the trained condition as well as an untrained observer speed condition. Results demonstrate that collision detection performance for older participants can be improved with perceptual learning and may transfer to untrained observer speed
Preferred Following Distance and Performance in an Emergency Event while Using Cooperative Adaptive Cruise Control
No full textThis study explored human factors issues associated with cooperative adaptive cruise control (CACC); specifically the relationship between drivers’ preferred following distance, assigned following distance, and driving performance. Participants drove in a dedicated lane and experienced a vehicle merging in front of their vehicle and later, an emergency event that required intervention in order to avoid a collision. Drivers followed at either a near or a far distance. Drivers’ perceived workload did not vary between the cruise and postmerge periods. However, workload was significantly greater after the emergency crash event. Workload did not vary significantly based on following distance assignment or preference. Those participants assigned to the near following distance were more likely to hover their foot over the brake during the merging event and to react faster to the emergency event. As with workload, performance (collision avoidance) did not vary significantly due to following distance assignment or preference. In other words, one’s abilities may not necessarily reflect their following preferences. This is a promising finding for widespread implementation of CACC