729 research outputs found
Sort by
Using a Driving Simulator to Create a Visual Search Test for Drivers with Parkinson's Disease
No full textVisual search has been reported as one of the most important determinants of on-road driving in Parkinson’s disease (PD). Yet, commonly used visual search tests are administered on paper or on a computer and have no to little face validity. This study aimed to (1) create a visual search test in a driving simulator; (2) investigate the convergent validity of the test against the dot cancellation (DC) test; and (3) compare performance on the test between 20 drivers with PD and 15 controls. Participants searched for a target road sign among a clutter of other road signs on three screens with 100° of horizontal visual field. Drivers with PD took longer to respond (9s ± 2 vs 7s ± 1; p = 0.001) and missed more target road signs (1.50 (0.5 – 7) vs 0 (0 – 1); p = 0.01) than controls. No differences were found between groups on the DC test. Response time on the visual search test correlated strongly with DC time (r = 0.52; p = 0.009) and moderately with DC errors (r = 0.37; p = 0.03). Missed responses correlated moderately with DC time (r = 0.49; p = 0.02). Our findings suggest that the driving simulator visual search test offers a valid alternative to standard visual search tests. Future research is needed to investigate the validity of the new visual search test in predicting on-road driving performance in PD
Drivers Fail to Calibrate to Optic Flow Speed Changes During Automated Driving
No full textThe human perceptual-motor system remains well-calibrated during manual driving supporting successful steering despite changing conditions, such as alterations in vehicle speed. Automated vehicles may interrupt perceptual-motor calibration so that when a driver takes-over control they will not be prepared for the driving conditions. Optic flow is a powerful source of visual information for calibrating to speed changes during manual steering, but it is currently unclear whether humans are sensitive to changes in optic flow speed when they are not in active control of the vehicle (i.e. by relying upon vision alone). Here we used a driving simulator to examine sensitivity to changes in optic flow speed across active (manual steering) and passive (automated steering) modes of control. Optic flow speed was altered independent of vehicle speed. The mismatch between perceived speed and actual speed causes a well-calibrated motor system to be reliably biased. Drivers were asked to take-over manual steering control after a short (~10 s) period of automation. Results showed that manual steering was not biased when flow speed was manipulated only in the automated period. One interpretation is that drivers had trouble recalibrating to optic flow changes that occurred during automated driving. If so, this suggests that there will exist a period where the perceptual-motor system is miscalibrated in the early stages of take-over after automated vehicle control
Effects of Inaccurate Gaze Behavior on Young Drivers’ Hazard Anticipation
No full textA previous study (Yamani et al., 2018) demonstrated that the administration of expert eye movement videos following hazard anticipation training can improve the proportion of latent hazards anticipated by young drivers compared to control conditions. The current driving simulator study sought to examine whether the improvements observed in the previous study were merely due to drivers’ exposure to videos of the simulated driving scenarios with expert eye movement overlays immediately prior to evaluation, or whether modeling the accuracy of eye movement behavior can lead participants to internalize hazard anticipation skills more effectively. In a between-subject design, 36 drivers (18-21 years) were assigned to one of three experimental conditions – training only, training plus expert eye movements or training plus novice eye movements. All participants navigated four unique driving scenarios, each with their eye movements tracked and recorded. Analyses of the eye movement data showed that young drivers who saw the expert eye movement (accurate) videos immediately following training anticipated a substantially greater proportion of latent hazards compared to the young drivers that saw novice eye movement (inaccurate) videos following training. The data provide some evidence that drivers were able to successfully map and incorporate correct hazard anticipation glance behavior into their mental models. The findings present some implications for the design and evaluation of eye movement-based training interventions
In the Context of Whole Trips: New Insights Into Driver Management of Attention and Tasks
No full textIt is becoming increasingly important to understand how drivers strategically manage tasks and thread attention across time, as they drive through varying situations and conditions -- and as they have the opportunity to delegate tasks to vehicle automation while taking up other tasks themselves. To develop an understanding of these higher-level driver behaviors requires a research focus on longer periods of driving -- even on “whole trip” driving. It may also require new tools and methods. Therefore, to explore insights and implications of a “whole trip” focus, data from 10 drivers were analyzed using methods tailored for identifying patterns within larger sequences of driving data than single-task epochs. The results are reported, discussed, and contrasted with more conventional approaches based on single-task epochs
Replicating Test Track Protocols in a Simulator; What Needs to be Matched?
No full textMany different experimental methods are used to evaluate driving performance as well as to evaluate the effectiveness of various vehicle safety systems but the results often do not match between different experimental approaches. This study aimed to determine the extent to which results can be matched between a driving simulator and a test track when carefully designed studies are used to replicate findings. This study collected simulator data on the National Advanced Driving Simulator (NADS) at the University of Iowa to replicate findings concerning Forward-Crash-Warning interface effectiveness at the Vehicle Research and Test Center (VRTC), East Liberty Ohio. The simulator used a virtual replica of the test track as well as a road course. Event choreography and scanning behavior were compared. Results indicate that results from the simulator were similar to those obtained on the test track. This indicates simulators can replicate findings for the test track and are a valuable tool. Careful experimental design is required to match the event choreography to insure an appropriate comparison. An exact match of the driving environment was not needed for this interface evaluation to obtain comparable results. The extent to which matching motion cues was not evaluated and may prove challenging in simulators without motion systems
Role of Psychological Effects of Real-Time Travel Information Provision on En Route Traveler Route Choice Decisions
No full textWhile real-time travel information can aid travelers to make informed decisions, it may increase the cognitive load in information perception and the complexity in decision-making process, especially when the information is from multiple sources. Under this circumstance, human factors-related aspects in information perception and the consequent psychological effects of the information play significant roles in traveler route choice decision-making. This study proposes a hybrid route choice model under the presence of real-time travel information, in which latent variables are employed to represent the psychological effects of information provision. Three dimensions of psychological effects – cognitive burden, cognitive decisiveness, and emotional relief – are assumed as latent psychological constructs in the proposed model. Data on traveler behavior and information perception are obtained through interactive driving simulator-based experiments. Estimation results will verify that the inclusion of the latent variables enhances the understanding of route choice decision-making under information provision
Voice-Controlled In-Vehicle Systems: Effects of Voice-Recognition Accuracy in the Presence of Background Noise
No full textThis paper presents initial findings from a driving simulator studyThis paper presents initial findings from a driving simulator studycomparing user responses to a noise-robust voice-controlled system while drivingto a noise-sensitive one in the presence of background noise. Twenty participantsinteracted with both noise-sensitive and noise-robust simulated voice-controlledinfotainment systems while driving under three background noise conditions (nonoise, music, and children). While both systems were viewed as useful andsatisfying, user acceptance was affected by background noise with the noisesensitivesystem, but not the noise-robust one. There was also no evidence that useracceptance was calibrated by having background noise as a context for varyinglevels of accuracy. No significant differences were observed between the twosystems in driving performance metrics analyzed (average speed, speed variability,and standard deviation of lane position), but the use of either system affecteddriving performance compared to baseline driving. A larger sample size at the endof this study along with the analysis of a larger set of performance metrics willprovide further insights
Crash Risk Analysis of Distracted Driving Behavior: Influence of Secondary Task Engagement and Driver Characteristics
No full textDistracted driving has long been acknowledged as one of the main contributors to crashes in the US. According to past studies, driving behavior proved to be influenced by the socioeconomic characteristics of drivers. However, only few studies attempted to quantify that influence. The study proposed a Crash Risk Index to estimate the crash risk associated with the socioeconomic characteristics of drivers and their tendency to experience distracted driving. The analysis is conducted using data from the SHRP 2 Naturalistic Driving Study (NDS). The proposed Crash Risk Index (CRI) is developed based on a grading system of three measures: the crash risk associated with performing secondary tasks during driving, the effect of socioeconomic attributes (e.g. Age) on the likelihood of engagement in secondary tasks, and the effect of specific categories within each socioeconomic attribute (e.g. Age>60) on the likelihood of engagement in secondary tasks. Logistic Regression analysis was performed on the secondary tasks, socioeconomic attributes, and the specific socioeconomic characteristics. The results identified the significant secondary tasks with high crash risk and the socioeconomic characteristics with significant effect on determining drivers’ involvement in secondary tasks among each tested parameter. These results were used to quantify the grading system measures and hence estimate the proposed CRI. This index indicates the relative crash risk associated with the socioeconomic characteristics of drivers and considering the possibility of engagement in secondary tasks. The proposed CRI and the associated grading system are plausible methods for estimating auto insurance premiums
Relationship Between Brain Activity and Real-Road Driving Behavior: A Vector-Based Whole-Brain Functional Near-Infrared Spectroscopy Study
No full textAutomobile driving requires multiple brain functions. However, the brain regions related to driving behavior are unknown. Therefore, we measured activity of the frontal, parietal and occipital lobes during driving using functional near-infrared spectroscopy (fNIRS). Cortical activation patterns were examined in relation to driving behaviors, such as steering motion, accelerator pedal motion, and speed control. Six healthy adults participated in the experiment. Cerebral oxygen exchange (COE) was calculated based on the oxyhemoglobin and deoxyhemoglobin concentrations measured by fNIRS. The COE and driving behavior data were collected every 1 m and averaged for all subjects. Functional NIRS data for all 98 channels were extracted using principal component analysis. Similarity between extracted components and driving behaviors were confirmed by |cosine similarity|>0.3. Among the factors with confirmed similarity, we identified brain regions with high principal component loading (|PCL|>0.4). Among the 16 COE factors extracted, COE factor 1 and factor 5 exhibited similarity with steering motion (cosine similarity: factor 1, -0.538; factor 5, 0.551). The PCLs of COE factor 1 and factor 5 were high in the frontal lobe (Brodmann areas [BAs] 9, 8, and 4/3) (PCL>0.8). COE factor 6 exhibited a similarity with accelerator pedal motion (cosine similarity: 0.369), and the PCL of COE factor 6 was highest in the parietal lobe (BA7) (PCL= -0.62). Speed control did not exhibit similarity with any COE factor. These findings will contribute to the selection of brain measurement areas when fNIRS is used for vehicle driving assessment
The Relation Between the Driver Behavior Questionnaire, Demographics, and Driving History
No full textThis paper presents an analysis of responses obtained on the Driver Behavior Questionnaire (DBQ) and self-reported history of the frequency of crashes, citations, and warnings in a sample of 562 drivers. The sample was closely balanced by gender and distributed in a broadly proportional manner across an age range of from 20 to 69 years. As has been previously reported, age and gender were found to be related to both DBQ scores and crash rates. The size and demographic distribution of the sample allowed an analysis to be run looking at the relationships of DBQ subscale scores with crashes, citations, and warnings, while controlling for age and gender. The results show that higher violation scores are positively associated with increases in self-reported crash and citation likelihoods; the less serious but apparently more common experience of receiving a warning for one’s driving behavior has a significant positive association with both violation and lapse scores. The extent to which these findings can be considered relevant to the overall driving population is enhanced from previous research given the sample size and age/gender balance