Glance Duration Research Articles

Dynamics of Driver's Gaze: Explorations in Behavior Modeling and Maneuver Prediction

The study and modeling of driver's gaze dynamics are important because, if and how the driver is monitoring the driving environment is vital for driver assistance in manual mode, for takeover requests in highly automated mode and for semantic perception of the surround in fully autonomous mode. We developed a machine vision-based framework to classify driver's gaze into context rich zones of interest and model driver's gaze behavior by representing gaze dynamics over a time period using gaze accumulation, glance duration, and glance frequencies. As a use case, we explore the driver's gaze dynamic patterns during maneuvers executed in freeway driving, namely, left lane change maneuver, right lane change maneuver, and lane keeping. It is shown that condensing gaze dynamics into durations and frequencies leads to recurring patterns based on driver activities. Furthermore, modeling these patterns show predictive powers in maneuver detection up to a few hundred milliseconds a priori .

The impact of an in-vehicle display on glance distribution in partially automated driving in an on-road experiment

One of the major challenges of designing an HMI for partially automated vehicles is the trade-off between a sufficient level of system information and avoidance of distracting the driver. This study aimed to investigate drivers’ glance behavior as an indicator of distraction when vehicle guidance is partially automated. Therefore, an on-road experiment was conducted comparing two versions of an in-vehicle display (during partially automated driving) and no display (during manual driving) on a heavy congested highway segment. The distribution of drivers’ total glance durations on the HMI showed that visual attention was shifted away from monitoring the central road scene towards looking at the in-vehicle display to a considerable extent. However, an analysis of the distribution of single glance durations supports the view that using partial automation and a respective HMI does not lead to a critical increase in distraction. Driving with a simplified version of the HMI had the potential to reduce glance duration on and thus potential distraction of the in-vehicle display.

Effects of Quantity and Size of Buttons of In-Vehicle Touch Screen on Drivers’ Eye Glance Behavior

ABSTRACTThis article investigates the effects of the quantity and size of touch screen buttons and the task-interleaving strategies on drives’ eye glance behavior. An experiment was conducted on a fixed-base driving simulator with 20 participants. The participants were asked to perform a button search-and-press task on an in-vehicle touch screen while driving. A full-factorial within-subject design was used with three button quantities (4, 8, and 15) and three button sizes (14 mm, 24 mm, and 33 mm). Although a normal distribution was often assumed for the eye glance data in previous studies, our results show that the total eyes-off-road time (TEORT) and glance durations are generally not normally distributed (positively skewed) even after a log transformation. The results show that the number of buttons has an increasing effect on task completion time, TEORT, and long (2+ s) glances. However, in general, no such differences were found for button sizes. Further analysis shows that long glances were strongly associated with drivers completing the task with a single glance. It seems to suggest that a major cause of long glances is that drivers are reluctant to switch the task back to driving at subtask boundaries that are probably associated with the high cost of interruption. These findings confirm the importance of task resumability for in-vehicle user interfaces and have implications that careful task analysis needs to be conducted in the context of multitasking. Certain subtask combinations, such as a visual search followed by pressing the search target, may discourage task interleaving and ultimately compromise driving safety.

Examining drivers' eye glance patterns during distracted driving: Insights from scanning randomness and glance transition matrix

IntroductionVisual attention to the driving environment is of great importance for road safety. Eye glance behavior has been used as an indicator of distracted driving. This study examined and quantified drivers' glance patterns and features during distracted driving. MethodData from an existing naturalistic driving study were used. Entropy rate was calculated and used to assess the randomness associated with drivers' scanning patterns. A glance-transition proportion matrix was defined to quantity visual search patterns transitioning among four main eye glance locations while driving (i.e., forward on-road, phone, mirrors and others). All measurements were calculated within a 5s time window under both cell phone and non-cell phone use conditions. ResultsResults of the glance data analyses showed different patterns between distracted and non-distracted driving, featured by a higher entropy rate value and highly biased attention transferring between forward and phone locations during distracted driving. Drivers in general had higher number of glance transitions, and their on-road glance duration was significantly shorter during distracted driving when compared to non-distracted driving. ConclusionsResults suggest that drivers have a higher scanning randomness/disorder level and shift their main attention from surrounding areas towards phone area when engaging in visual-manual tasks. Practical applicationsDrivers' visual search patterns during visual-manual distraction with a high scanning randomness and a high proportion of eye glance transitions towards the location of the phone provide insight into driver distraction detection. This will help to inform the design of in-vehicle human-machine interface/systems.

Revisiting Radio Tuning: A Secondary Analysis Comparing Glance Behavior Across Five Vehicles

Visual-manual demand placed on drivers through interactions with operational functions, embedded telematics, infotainment systems, and nomadic technologies has raised concerns associated with diverting attention from the roadway. This analysis draws on data from field studies of five different infotainment systems representing a range of screen placements and control characteristics that diverge in significant ways from the relatively standard car radio layout of previous generations. Participants performed a set of classic visual-manual radio tuning tasks under highway driving conditions. There were significant differences in task completion time, number of off-road glances, mean single off-road glance duration, and total off-road glance time across vehicles. These results highlight that the range of configurations appearing in modern infotainment systems have changed the extent to which they can be used in the classic radio tuning task to provide a standard demand benchmark.

Effects of written peer-feedback content and sender’s competence on perceptions, performance, and mindful cognitive processing

Peer-feedback efficiency might be influenced by the oftentimes voiced concern of students that they perceive their peers’ competence to provide feedback as inadequate. Feedback literature also identifies mindful processing of (peer)feedback and (peer)feedback content as important for its efficiency, but lacks systematic investigation. In a 2 × 2 factorial design, peer-feedback content (concise general feedback [CGF] vs. elaborated specific feedback [ESF]) and competence of the sender (high vs. low) were varied. Students received a scenario containing an essay by a fictional student and fictional peer feedback, a perception questionnaire, and a text revision, distraction, and peer-feedback recall task. Eye tracking was applied to measure how written peer feedback was (re-)read, e.g., glance duration on exact words and sentences. Mindful cognitive processing was inferred from the relation between glance duration and (a) text-revision performance and (b) peer-feedback recall performance. Feedback by a high competent peer was perceived as more adequate. Compared to CGF, participants who received ESF scored higher on positive affect towards the peer feedback. No effects were found for peer-feedback content and/or sender’s competence level on performance. Glance durations were negatively correlated to text-revision performance regardless of condition, although peer-feedback recall showed that a basic amount of mindful cognitive processing occurred in all conditions. Descriptive findings also hint that this processing might be dependent on an interaction between peer-feedback content and sender’s competence, signifying a clear direction for future research.

Glass half-full: On-road glance metrics differentiate crashes from near-crashes in the 100-Car data

BackgroundMuch of the driver distraction and inattention work to date has focused on concerns over drivers removing their eyes from the forward roadway to perform non-driving-related tasks, and its demonstrable link to safety consequences when these glances are timed at inopportune moments. This extensive literature has established, through the analyses of glance from naturalistic datasets, a clear relationship between eyes-off-road, lead vehicle closing kinematics, and near-crash/crash involvement. ObjectiveThis paper looks at the role of driver expectation in influencing drivers’ decisions about when and for how long to remove their eyes from the forward roadway in an analysis that consider the combined role of on- and off-road glances. MethodUsing glance data collected in the 100-Car Naturalistic Driving Study (NDS), near-crashes were examined separately from crashes to examine how momentary differences in glance allocation over the 25-s prior to a precipitating event can differentiate between these two distinct outcomes. Individual glance metrics of mean single glance duration (MSGD), total glance time (TGT), and glance count for off-road and on-road glance locations were analyzed. Output from the AttenD algorithm (Kircher and Ahlström, 2009) was also analyzed as a hybrid measure; in threading together on- and off-road glances over time, its output produces a pattern of glance behavior meaningful for examining attentional effects. ResultsIndividual glance metrics calculated at the epoch-level and binned by 10-s units of time across the available epoch lengths revealed that drivers in near-crashes have significantly longer on-road glances, and look less frequently between on- and off- road locations in the moments preceding a precipitating event as compared to crashes. During on-road glances, drivers in near-crashes were found to more frequently sample peripheral regions of the roadway than drivers in crashes. Output from the AttenD algorithm affirmed the cumulative net benefit of longer on-road glances and of improved attention management between on- and off-road locations. ConclusionThe finding of longer on-road glances differentiating between safety-critical outcomes in the 100-Car NDS data underscores the importance of attention management in how drivers look both on and off the road. It is in the pattern of glances to and from the forward roadway that drivers obtained critical information necessary to inform their expectation of hazard potential to avoid a crash. ApplicationThis work may have important implications for attention management in the context of the increasing prevalence of in-vehicle demands as well as of vehicle automation.

Driver performance and attention allocation in use of logo signs on freeway exit ramps

The objective of this research was to quantify the effects of driver age, ramp signage configuration, including number of panels, logo format and sign familiarity, on driver performance and attention allocation when exiting freeways. Sixty drivers participated in a simulator study and analysis of variance models were used to assess response effects of the controlled manipulations. Results revealed elderly drivers to demonstrate worse performance and conservative control strategies as compared to middle-aged and young drivers. Elderly drivers also exhibited lower off-road fixation frequency and shorter off-road glance durations compared to middle-aged and young drivers. In general, drivers adopted a more conservative strategy when exposed to nine-panel signs as compared to six-panel signs and were more accurate in target detection when searching six-panels vs. nine and with familiar vs. unfamiliar logos. These findings provide an applicable guide for agency design of freeway ramp signage accounting for driver demographics.

Designing feedback to mitigate teen distracted driving: A social norms approach

ObjectiveThe purpose of this research is to investigate teens’ perceived social norms and whether providing normative information can reduce distracted driving behaviors among them. BackgroundParents are among the most important social referents for teens; they have significant influences on teens’ driving behaviors, including distracted driving which significantly contributes to teens’ crash risks. Social norms interventions have been successfully applied in various domains including driving; however, this approach is yet to be explored for mitigating driver distraction among teens. MethodForty teens completed a driving simulator experiment while performing a self-paced visual-manual secondary task in four between-subject conditions: a) social norms feedback that provided a report at the end of each drive on teens’ distracted driving behavior, comparing their distraction engagement to their parent’s, b) post-drive feedback that provided just the report on teens’ distracted driving behavior without information on their parents, c) real-time feedback in the form of auditory warnings based on eyes of road-time, and d) no feedback as control. Questionnaires were administered to collect data on these teens’ and their parents’ self-reported engagement in driver distractions and the associated social norms. ResultsSocial norms and real-time feedback conditions resulted in significantly smaller average off-road glance duration, rate of long (>2s) off-road glances, and standard deviation of lane position compared to no feedback. Further, social norms feedback decreased brake response time and percentage of time not looking at the road compared to no feedback. No major effect was observed for post-drive feedback. Questionnaire results suggest that teens appeared to overestimate parental norms, but no effect of feedback was found on their perceptions. ConclusionFeedback systems that leverage social norms can help mitigate driver distraction among teens. Overall, both social norms and real-time feedback induced positive driving behaviors, with social norms feedback outperforming real-time feedback.

Developing Predictive Equations to Model the Visual Demand of In-Vehicle Touchscreen HMIs

ABSTRACTTouchscreen human–machine interfaces (HMIs) are commonly employed as the primary control interface and touch-point of vehicles. However, there has been very little theoretical work to model the demand associated with such devices in the automotive domain. Instead, touchscreen HMIs intended for deployment within vehicles tend to undergo time-consuming and expensive empirical testing and user trials, typically requiring fully functioning prototypes, test rigs, and extensive experimental protocols. While such testing is invaluable and must remain within the normal design/development cycle, there are clear benefits, both fiscal and practical, to the theoretical modeling of human performance. We describe the development of a preliminary model of human performance that makes a priori predictions of the visual demand (total glance time, number of glances, and mean glance duration) elicited by in-vehicle touchscreen HMI designs, when used concurrently with driving. The model incorporates information theoretic components based on Hick–Hyman Law decision/search time and Fitts’ Law pointing time and considers anticipation afforded by structuring and repeated exposure to an interface. Encouraging validation results, obtained by applying the model to a real-world prototype touchscreen HMI, suggest that it may provide an effective design and evaluation tool, capable of making valuable predictions regarding the limits of visual demand/performance associated with in-vehicle HMIs, much earlier in the design cycle than traditional design evaluation techniques. Further validation work is required to explore the behavior associated with more complex tasks requiring multiple screen interactions, as well as other HMI design elements and interaction techniques. Results are discussed in the context of facilitating the design of in-vehicle touchscreen HMI to minimize visual demand.

Smartwatches vs. Smartphones

This work seeks to understand whether the unique features of a smartwatch, compared to a smartphone, mitigate or exacerbate driver distraction due to notifications, and to provide insights about drivers' perceptions of the risks associated with using smartwatches while driving. As smartwatches are gaining popularity among consumers, there is a need to understand how smartwatch use may influence driving performance. Previous driving research has examined voice calling on smartwatches, but not interactions with notifications, a key marketed feature. Engaging with notifications (e.g., reading and texting) on a handheld device is a known distraction associated with increased crash risks. Two driving simulator studies compared smartwatch to smartphone notifications. Experiment I asked participants to read aloud brief text notifications and Experiment II had participants manually select a response to arithmetic questions presented as notifications. Both experiments investigated the resulting glances to and physical interactions with the devices, as well as self-reported risk perception. Experiment II also investigated driving performance and self-reported knowledge/expectation about legislation surrounding the use of smart devices while driving. Experiment I found that participants were faster to visually engage with the notification on the smartwatch than the smartphone, took longer to finish reading aloud the notifications, and exhibited more glances longer than 1.6 s. Experiment II found that participants took longer to reply to notifications and had longer overall glance durations on the smartwatch than the smartphone, along with longer brake reaction times to lead vehicle braking events. Compared to the no device baseline, both devices increased lane position variability and resulted in higher self-reported perceived risk. Experiment II participants also considered that smartwatch use while driving deserves penalties equal to or less than smartphone use. The findings suggest that smartwatches may have road safety consequences. Given the common view among participants to associate smartwatch use with equal or less traffic penalties than smartphone use, there may be a disconnect between drivers' actual performance and their perceptions about smartwatch use while driving.

Assessing drivers' response during automated driver support system failures with non-driving tasks

IntroductionWith the increase in automated driver support systems, drivers are shifting from operating their vehicles to supervising their automation. As a result, it is important to understand how drivers interact with these automated systems and evaluate their effect on driver responses to safety critical events. This study aimed to identify how drivers responded when experiencing a safety critical event in automated vehicles while also engaged in non-driving tasks. MethodIn total 48 participants were included in this driving simulator study with two levels of automated driving: (a) driving with no automation and (b) driving with adaptive cruise control (ACC) and lane keeping (LK) systems engaged; and also two levels of a non-driving task (a) watching a movie or (b) no non-driving task. In addition to driving performance measures, non-driving task performance and the mean glance duration for the non-driving task were compared between the two levels of automated driving. ResultsDrivers using the automated systems responded worse than those manually driving in terms of reaction time, lane departure duration, and maximum steering wheel angle to an induced lane departure event. These results also found that non-driving tasks further impaired driver responses to a safety critical event in the automated system condition. ConclusionIn the automated driving condition, driver responses to the safety critical events were slower, especially when engaged in a non-driving task. Practical applicationTraditional driver performance variables may not necessarily effectively and accurately evaluate driver responses to events when supervising autonomous vehicle systems. Thus, it is important to develop and use appropriate variables to quantify drivers' performance under these conditions.

Evaluation of the effects of in‐vehicle traffic lights on driving performances for unsignalised intersections

Ground traffic lights are essential for maintaining traffic efficiency and safety at intersections. However, unsignallised intersections are still frequent in actual traffic environments. With the development of new forms of vehicular communication, in-vehicle traffic lights can assist drivers at unsignallised intersections. The authors proposed two types of in-vehicle traffic lights to assist drivers; these corresponded to two-way and all-way stop-controlled intersections. They adopted gap acceptance theory and a first-come-first-served strategy to determine passing priority for the two types of intersections, respectively. They then conducted a driving simulator experiment involving 23 participants, to investigate driver behaviours elicited by the proposed system. They prepared four experimental conditions with combinations of in-vehicle traffic lights and auditory warnings. The authors’ experimental results indicated that in-vehicle traffic lights were associated with significantly longer post-encroachment times and a significantly shorter maximum brake stroke. In terms of eye-gaze behaviours, the percentage of gaze concentration to the road centre area and mean glance durations were deemed acceptable for the avoidance of visual distraction, when in-vehicle traffic lights were presented via a head-up display. Therefore, their analysis of driver behaviours indicates that in-vehicle traffic lights can effectively provide driver assistance at unsignallised intersections.

Linking the Detection Response Task and the AttenD Algorithm Through Assessment of Human–Machine Interface Workload

Multitasking related demands can adversely affect drivers’ allocation of attention to the roadway, resulting in delays or missed responses to roadway threats and to decrements in driving performance. Robust methods for obtaining evidence and data about demands on and decrements in the allocation of driver attention are needed as input for design, training, and policy. The detection response task (DRT) is a commonly used method (ISO 17488) for measuring the attentional effects of cognitive load. The AttenD algorithm is a method intended to measure driver distraction through real-time glance analysis, in which individual glances are converted into a scalar value using simple rules considering glance duration, frequency, and location. A relationship between the two tools is explored. A previous multitasking driving simulation study, which used the remote form of the DRT to differentiate the demands of a primary visual–manual human–machine interface from alternative primary auditory–vocal multimodal human–machine interfaces, was reanalyzed using AttenD, and the two analyses compared. Results support an association between DRT performance and AttenD algorithm output. Summary statistics produced from AttenD profiles differentiate between the demands of the human–machine interfaces considered with more power than analyses of DRT response time and miss rate. Among discussed implications is the possibility that AttenD taps some of the same attentional effects as the DRT. Future research paths, strategies for analyses of past and future data sets, and possible application for driver state detection are also discussed.

On-Road Visual Sign Salience, Driver Attention Allocation, and Target Detection Accuracy

Previous studies have empirically assessed the effect of on-road signage on driver distraction and attention allocation through eye-tracking measures and driving simulations. However, state and federal agencies might be able to make use of simpler approaches to determining the potential visual load of on-road signs. This study compared different on-road signage (food, gas and attraction, and distance guide signs) with respect to visual salience, assessed the effect of the number of panels and logo format on signage salience, and analyzed associations between salience values for different sign types with target detection accuracy and off-road glance duration. Visual salience of signs was measured with the salience using the natural statistics (SUNs) method. For correlation analyses between sign salience, attention allocation, and detection accuracy, a sample of 120 drivers participated in two driving simulation studies. Results revealed food business signs, containing all pictorial logos, to be more salient than gas and attraction and distance guide signs. In addition, results revealed that increasing the number of sign panels from six to nine significantly increased visual salience. It was also found that attraction signs including all pictorial logos were more salient than signs with text-based logos. Results did not reveal statistically significant relationships between visual salience and attention allocation. In addition, there was no significant correlation between target detection accuracy and visual salience for various sign types. Future work with a higher sign-sampling rate is needed to assess whether the SUNs method is a useful measure for determining if roadway sign salience is a driving factor in distraction.

Modeling Effects of Forward Glance Durations on Latent Hazard Detection

Performance of in-vehicle, secondary tasks requires a driver to alternate his or her glances between the inside of the vehicle and the forward roadway. While previous research has shown that thresholds of off-road and durations are critical to latent hazard detection, there has been no research to predict the probability of hazard detection in a time series considering all possible forward glance durations. To determine the minimum forward roadway duration to adequately anticipate hazards, 45 drivers were asked to navigate eight unique virtual scenarios on a driving simulator while alternating their glances inside (2s) and outside (1s, 2s, 3s, or 4s) the vehicle in a uniform sequence (consistent on-road and off-road durations). A microbenchmark approach based on Hidden Markov models is explored to infer the transition probability of hazard detection that changes dynamically between the detection and nondetection stages. The model is cross validated and demonstrated to be accurate and robust. Three characteristics of total sequence time were tested in the model. Using the ground truth transition probability from fixed duration, the probability of hazard detection with variable duration within an alternation sequence was computed. Across nonuniform alternation sequences, different permutations of sequences show that a short time series of alternation (10s) of glance behavior is sufficient for hazard detection (greater than 50.0%). Appropriate countermeasures can be developed to increase a driver’s forward glance duration whenever the detection probability is predicted low.

Optimising landmark-based route guidance for older drivers

In-vehicle navigation systems (IVNS) have the potential to benefit older drivers, reducing stress associated with way-finding and providing on-trip support, especially in unfamiliar locations. However, existing IVNS present challenges to usability, resulting in lack of uptake and over-reliance on pre-trip planning.This paper presents research aimed at identifying features that make IVNS user-friendly and appropriate for older drivers. Studying navigational performance within a simulated driving environment, it focuses on the use of landmarks with route guidance information, and the most appropriate method of information provision (audio only, visual only or a combination of audio and visual). It also assesses potential gender differences that might arise with landmark-based navigational information.Solutions include use of appropriate roadside landmarks, and information delivered through a combination of audio and icon-based visual format. These features result in lower workload and fewer navigational errors. The audio/visual modality reduces the hazard of distraction by landmarks resulting in fewer visual glances and lower glance duration to the roadside compared to other modalities.Design and provision of IVNS tailored to older drivers’ needs can make a considerable contribution to maintaining individual mobility for longer.

Drivers' Visual Characteristics when Merging onto or Exiting an Urban Expressway.

The aim of this study is to examine drivers’ visual and driving behavior while merging onto or exiting an urban expressway with low and high traffic densities. The analysis was conducted according to three periods (approaching, merging or exiting, and accelerating or decelerating). A total of 10 subjects (8 males and 2 females) with ages ranging from 25 to 52 years old (M = 30.0 years old) participated in the study. The research was conducted in a natural driving situation, and the drivers’ eye movements were monitored and recorded using an eye tracking system. The results show that the influence of traffic density on the glance duration and scan duration is more significant when merging than when exiting. The results also demonstrate that the number of glances and the mean glance duration are mainly related to the driving task (e.g., the merging period). Therefore, drivers’ visual search strategies mainly depend on the current driving task. With regard to driving behavior, the variation tendencies of the duration and the velocity of each period are similar. These results support building an automated driving assistant system that can automatically identify gaps and accelerate or decelerate the car accordingly or provide suggestions to the driver to do so.

Sensation Seeking and Drivers’ Glance Behavior while Engaging in a Secondary Task

To investigate possible relationships between drivers’ sensation seeking and glance behavior while interacting with human-machine interfaces, a total of 70 drivers’ eye-glance data, Sensation Seeking Scale (SSS), and Driver Behavior Questionnaire (DBQ) data were collected and analyzed. Participants conducted radio tuning tasks with two standard production interfaces while driving on a highway, and their glance allocations to defined regions were recorded and manually annotated. Results showed that sensation seeking scores were related with self-reported violation scores, off-road glance patterns, and driving speed: (1) violation scores of DBQ were positively correlated with sensation seeking, (2) mean and standard deviation of off-road glance duration were positively correlated with sensation seeking for younger drivers (under 40 years), (3) total off-road glance time per minute and number of off-road glances per minute were positively correlated with sensation seeking for older drivers (over 40 years), and (4) percentage of speed change was negatively correlated with sensation seeking for both younger and older drivers. The results indicate that sensation seeking is associated with drivers’ off-road glance patterns and driving behavior. These observations further highlight the relationship between personal traits and driver behavior.

Gaming to Safety

The risk of crash or near-crash significantly increases when drivers make long off-road glances to engage in distracting tasks. Providing drivers with feedback that integrates elements of game design could increase driver motivation for adopting safer behaviors. In an ongoing between-subjects simulator study with young drivers ( n=29 reported in this paper), we compare four conditions for off-road glance behaviors: no feedback, real-time feedback system, post-drive feedback system (real-time feedback + post-drive feedback), and gamification feedback system (real-time feedback + post-drive feedback + game design elements). Shorter average glance duration and less frequent risky (≥2 s) glances to an in-vehicle display were observed for the post-drive system, compared to no feedback and real-time feedback, and for the gamification system, compared to no feedback. Although no added benefit of gamification over the post-drive feedback system was observed for these eye glance metrics, longer-term exposure and assessment could show improvements to be more stable with the inclusion of game design elements.