Triangulated Investigation of Trust in Automated Driving: Challenges and Solution Approaches for Data Integration
In automated driving, an appropriate level of driver trust is essential to improve safety and ensure zero fatalities. Drivers must have a sufficient level of trust to intervene correctly in safety-critical situations: very low levels may lead to either continuous and excessive monitoring of the functions, reducing the attention paid to the environment or switching off these functions, whereas extreme trust in automated driving functions can result in dangerous driving situations because the environment is either insufficiently monitored or not monitored at all. A deeper understanding of trust in automated driving is challenging and requires a triangulated study in which the type of driver, vehicle usage, and environmental data are varied. However, many previous studies were based on a rather limited set of data sources, often relying on qualitative means such as pre-and-post interviews or trust questionnaires to evaluate trust in autonomous driving functions. Although data gathered through empirical research, such as conducting quantitative surveys or qualitative interviews, are simple to store and analyze, the collection and integration of vehicle and sensor data from different data sources usually pose important technical challenges in practice. Hence, a suitable data collection and integration strategy is required to address these challenges. In this context, we propose a general framework for collecting and integrating data from different sources with diverse capabilities and requirements to determine a driver’s trust in automated driving. Our proposed framework facilitates the integration of empirical and measurement data, allowing a triangulated investigation to provide a road map for the automotive industry.
Eye Tracking Glasses
Software
Validity of primary driving tasks in head-mounted display-based driving simulators
The development of new car interior concepts requires tools, particularly in development phases before concept milestones, which enable subjective experiences and evaluations in static and driving situations. On the one hand, variant comparisons are necessary; on the other hand, the level of immersion should be high enough that participants can behave as they would in real cars. Virtual reality technologies and especially head-mounted displays are generally very suitable for such evaluations with the exception being in state-of-the-art driving simulators. Therefore, a validation study was undertaken in which primary driving tasks in two HMD-based simulators were compared with test runs in a real car. The difference between the simulators was only the state of the motion base (enabled vs. disabled). In both simulators and the test runs in the real car, four identical scenarios (straight, curves, overtaking and junction) were carried out. Since the focus is primarily on subjective ratings and gaze behaviour when evaluating new car interior concepts, in this study gaze behaviour was also priority. In addition, driving dynamics parameters were measured. The results reveal that the participants show more valid behaviour in the dynamic system than in the static simulator condition.
Visual enhancements for the driver’s information search on automotive head-up display
In the past, in-vehicle head-up displays (HUDs) were used to display simple information including driving speed and the distance between cars. However, recent HUDs now display complex information such as advanced driver assistance information. This study aims to identify the effects of visual enhancements for HUDs on the driver’s performance and workload. Twenty participants conducted the tracking tasks for information search while driving in an automotive simulator environment. The participants experienced three levels of visual enhancements (none, shaded reference bar, translucent reference bar) for each task difficulty (low, medium, high). The results showed that visual enhancements and task difficulties had a significant effect on the tracking errors and subjective workloads. These findings verify that the translucent reference bar significantly improved the tracking performance. Furthermore, the visual enhancement cues on the HUDs play an important role in visual search. This research provides practical guidelines to ensure road safety through minimizing cognitive workload on drivers. Therefore, the results will encourage interface designers to consider the visual enhancement for HUDs from a user-centered perspective.
Evaluation of ambient light displays for requests to intervene and minimal risk maneuvers in highly automated urban driving
Driver take-overs at system limits and the corresponding requests to intervene are not obligatory in highly automated driving. Therefore, minimal risk maneuvers may occur. In order to clearly communicate the automation status and the driver's task at such system limits, ambient light displays seem to have a high potential. Therefore, two ambient light displays were investigated in a driving simulation experiment, mounted either at the bottom of the windshield or on the steering wheel. Forty participants experienced two request to intervene scenarios, and two minimal risk maneuver scenarios during a highly automated drive. In general, both ambient light displays seem to ensure correct driver's reactions and safety, and lead to positive subjective ratings. Results revealed no significant differences between the two ambient light display positions regarding the take-over behavior, gaze behavior and subjective rating, except in terms of the perceived brightness.
Eye Tracking Glasses
Simulator
Gender aspects of trust in and use of advanced driver assistance systems/submitted by Gina Nathalie Schnücker
Increasing automation in vehicles promises an equal increase in driving safety and comfort. Current research aims to identify factors that influence if and how in-vehicle automation is adopted by the human operator. Trust in automation was identified to be one of those factors. Still, there is a need to understand the relationship between system-trust and actual reliance behavior. Therefore, the first aim of this thesis was to investigate the relationship between trust in Advanced Driver Assistance Systems and actual reliance behavior. The second aim was to address the inconsistencies in previous research regarding gender differences in automation trust by introducing the concept of gender roles. The concept of gender roles addresses stereotypical behavior associated with masculinity and femininity rather than biological sex and has already been identified to explain variance in addition and beyond biological sex in manual driving, risk-taking, and technology adoption. Qualitative, quantitative, and psychophysiological data were integrated in an experiment with 100 participants (49 females) in a 25-minute field-drive. We assessed attitudes and behavior in interaction with a Lane Keeping Assist (LA), that supports the driver to keep the vehicle within the lane and the Adaptive Cruise Control (ACC) that maintains a set speed and keeps a set distance to preceding vehicles. The results show that trust after the automated drive was positively related to the use of the LA and the simultaneous use of LA and ACC. It also showed that there were no sex differences in trust and reliance behavior. In contrast, masculinity was positively related to trust before the automated drive, while femininity was positively related to trust after the drive. These findings have implications for trust measurement and the promotion of a safe adoption of in-vehicle automation.
Eye Tracking Glasses
Simulator
Human Factors Studies on Automotive Head-Up Display Design
Human Factors Studies on Automotive Head-Up Display Design Juhee Park Department of Industrial Engineering The Graduate School Seoul National University Head-up display (HUD) systems were introduced into the automobile industry as a means for improving driving safety. They superimpose safety-critical information on top of the driver’s forward field of view and thereby help drivers keep their eyes forward while driving. Since the first introduction about three decades ago, automotive HUDs have been available in various commercial vehicles. Despite the long history and potential benefits of automotive HUDs, however, the design of useful automotive HUDs remains a challenging problem. In an effort to contribute to the design of useful automotive HUDs, this doctoral dissertation research conducted four studies. In Study 1, the functional requirements of automotive HUDs were investigated by reviewing the major automakers' automotive HUD products, academic research studies that proposed various automotive HUD functions, and previous research studies that surveyed drivers’ HUD information needs. The review results indicated that: 1) the existing commercial HUDs perform largely the same functions as the conventional in-vehicle displays, 2) past research studies proposed various HUD functions for improving driver situation awareness and driving safety, 3) autonomous driving and other new technologies are giving rise to new HUD information, and 4) little research is currently available on HUD users’ perceived information needs. Based on the review results, this study provides insights into the functional requirements of automotive HUDs and also suggests some future research directions for automotive HUD design. In Study 2, the interface design of automotive HUDs for communicating safety-related information was examined by reviewing the existing commercial HUDs and display concepts proposed by academic research studies. Each display was analyzed in terms of its functions, behaviors and structure. Also, related human factors display design principles, and, empirical findings on the effects of interface design decisions were reviewed when information was available. The results indicated that: 1) information characteristics suitable for the contact-analog and unregistered display formats, respectively, are still largely unknown, 2) new types of displays could be developed by combining or mixing existing displays or display elements at both the information and interface element levels, and 3) the human factors display principles need to be used properly according to the situation and only to the extent that the resulting display respects the limitations of the human information processing, and achieving balance among the principles is important to an effective design. On the basis of the review results, this review suggests design possibilities and future research directions on the interface design of safety-related automotive HUD systems. In Study 3, automotive HUD-based take-over request (TOR) displays were developed and evaluated in terms of drivers’ take-over performance and visual scanning behavior in a highly automated driving situation. Four different types of TOR displays were comparatively evaluated through a driving simulator study - they were: Baseline (an auditory beeping alert), Mini-map, Arrow, and Mini-map-and-Arrow. Baseline simply alerts an imminent take-over, and was always included when the other three displays were provided. Mini-map provides situational information. Arrow presents the action direction information for the take-over. Mini-map-and-Arrow provides the action direction together with the relevant situational information. This study also investigated the relationship between driver’s initial trust in the TOR displays and take-over and visual scanning behavior. The results indicated that providing a combination of machine-made decision and situational information, such as Mini-map-and-Arrow, yielded the best results overall in the take-over scenario. Also, drivers’ initial trust in the TOR displays was found to have significant associations with the take-over and visual behavior of drivers. The higher trust group primarily relied on the proposed TOR displays, while the lower trust group tended to more check the situational information through the traditional displays, such as side-view or rear-view mirrors. In Study 4, the effect of interactive HUD imagery location on driving and secondary task performance, driver distraction, preference, and workload associated with use of scrolling list while driving were investigated. A total of nine HUD imagery locations of full-windshield were examined through a driving simulator study. The results indicated the HUD imagery location affected all the dependent measures, that is, driving and task performance, drivers’ visual distraction, preference and workload. Considering both objective and subjective evaluations, interactive HUDs should be placed near the driver's line of sight, especially near the left-bottom on the windshield.
Eye Tracking Glasses
Simulator
Impact of urban undersea tunnel longitudinal slope on the visual characteristics of drivers
This study aims to investigate the impact of the urban undersea tunnel longitudinal slope on the visual characteristics of drivers. 20 drivers were enrolled to conduct the real vehicle test of the urban undersea tunnel. First, the data of average fixation time and visual lobe were collected by an eye tracker. The differential significance was tested using the one-way repeated measures analysis of variance (ANOVA). Then, the difference between the up-and-down slope (direction) factor and the longitudinal slope (percent) factor on the two indexes were analyzed using the two-way repeated measures ANOVA. Second, by constructing a Lorentz model, the impact of the longitudinal slope on the average fixation time and the visual lobe were analyzed. Besides, a three-dimensional model of the longitudinal slope, average fixation time, and visual lobe was quantified. The results showed that the average fixation time and visual lobe under different longitudinal slopes markedly differed when driving on the uphill and downhill sections. The average fixation time and visual lobe under two factors were markedly different. Moreover, with an increase in the longitudinal slope, the average fixation time exhibited a trend of increasing first then decreasing; the visual lobe exhibited a trend of decreasing first and then increasing. The average fixation time reached the minimum and maximum value when the slope was 2.15% and 4.0%, whereas the visual lobe reached the maximum and minimum value when the slope was 2.88% and 4.0%. Overall, the longitudinal slope exerted a great impact on the visual load of the driver.
Eye Tracking Glasses
Simulator
Multimodal features for detection of driver stress and fatigue
Driver fatigue and stress significantly contribute to higher number of car accidents worldwide. Although, different detection approaches have been already commercialized and used by car producers (and third party companies), research activities in this field are still needed in order to increase the reliability of these alert systems. Also, in the context of automated driving, the driver mental state assessment will be an important part of cars in future. This paper presents state-of-the-art review of different approaches for driver fatigue and stress detection and evaluation. We describe in details various signals (biological, car and video) and derived features used for these tasks and we discuss their relevance and advantages. In order to make this review complete, we also describe different datasets, acquisition systems and experiment scenarios.
Eye Tracking Glasses
Simulator
Research on drivers’ visual characteristics in different curvatures and turning conditions of the extra-long urban underwater tunnels
In order to study driver’s visual characteristics under different curvatures and turning conditions in extra-long urban underwater tunnels, fixation and saccade were herein regarded as the main research objectives. In this study, we carried out real vehicle testing on curved sections with 5 different radii and straight sections of the extra-long urban underwater tunnels. The driver’s fixation characteristics were studied by using fixation distribution, fixation time, fixation frequency, fixation time ratio, and frequency ratio. The driver’s saccade characteristics were investigated by selecting the saccade angle, saccade time, saccade frequency, saccade time ratio, and frequency ratio. Accordingly, mathematical models of the driver’s fixation time, fixation frequency, saccade time, and saccade frequency under different curvatures and turning conditions in the extra-long urban underwater tunnel were established. Combined with the change of visual distance, sight distance, and sight zone, driver’s visual characteristics in the extra-long urban underwater tunnel were further analyzed. The results demonstrated that the smaller the radius of the tunnel, the more focused driver’s fixation time, the greater the psychological pressure, and the lower the safety when driving. Under the same radius, driver’s tension and risk factors were higher during turning left, while driver’s driving mentality was more relaxed and driving situation was further stable in the right-turn.
Eye Tracking Glasses
Simulator
