Visualizing the autonomous vehicle’s maneuvers – Does an ecological interface help to increase the hedonic quality and safety?

This paper examines whether ecological speed information describing ongoing driving maneuvers during automated driving enhances the hedonic quality and driving safety immediately after a driving takeover. Visualizing maneuvers and trajectories has already proven effective. However, planned acceleration and deceleration in an automated vehicle have not yet been investigated. Therefore, this paper assesses how an automated vehicle’s speed control information might be presented by an ecological interface. Besides a possible increase in the hedonic quality, this information might enhance safe behavior of the human driver when it comes to a takeover. To assess these two aspects, 43 drivers participated in a dynamic driving simulator study. Using a within-subject design, two scenarios were used to compare an ecological interface, dynamically visualizing speed changes, to a conventional pop-up interface, using pop-up icons to visualize speed changes. The experimental results indicate that ecological feedback and conventional pop-up feedback do not differ regarding the hedonic quality, which was reflected by the state anxiety, usefulness, and satisfaction with the overall human-machine interface (HMI). Nonetheless, the post-hoc questionnaire on situational awareness showed a significantly lower rating for the ecological interface which may be the result of a more automatic and subconscious processing of the information given. Analyzing the takeover performance, the initial takeover time was comparably low for both interfaces. However, concerning safety, the ecological interface significantly enhanced the lateral control after takeover, and the drivers looked at the vehicle mirrors significantly earlier. In conclusion, the results show that the information given by the ecological interface may help drivers cope with a sudden takeover in a faster and more controlled way. Future applications of these findings might serve to enhance the acceptance and safety of semi-autonomous vehicles by implementing ecological interfaces.     

It’s about time! Earlier take-over requests in automated driving enable safer responses to conflicts

Automated driving (AD), which takes full responsibility for the driving task in certain conditions, is currently being developed. An important concern in AD is how to design a take-over request (TOR) that mitigates automation effects (e.g., delayed responses to conflict scenarios) that previous literature from simulator experiments has shown can occur. To address this concern, this study aims to investigate and compare driver responses to TORs and a lead-vehicle cut-out scenario under three conditions: (1) after a period of AD with a TOR issued early (18 s time-to-collision), (2) same as (1) except with a TOR issued late (9 s time-to-collision), and (3) baseline, with adaptive cruise control (ACC). This paper also compares the results to those of a previous study using the same conflict scenario but with near-perfect assisted driving system (SAE Level 2). The lead-vehicle cut-out scenario was encountered on a test track after 30 minutes driving with either ACC or AD. In AD the TOR was issued prior to the conflict object was revealed to the participants when the lead vehicle performed the cut-out (at conflict onset). This TOR strategy differed from previous driving-simulator studies that issued the TOR at conflict onset. The participants had to respond by steering and/or braking to avoid a crash. Our findings show that, independent of TOR timing, the drivers required similar amounts of time to 1) direct their first glance to the human–machine interface, 2) look forward, 3) end their secondary task, 4) put their hands on the steering wheel, and 5) deactivate automation. However, when the TOR was issued early rather than late, they started to brake earlier (even before conflict onset). All participants successfully managed to avoid crashing with the object, independent of the condition. AD with an early TOR resulted in the earliest response, while ACC drivers responded slightly earlier than the drivers in AD with the late TOR. Our findings do not support the findings of severe automation effects in previous driving-simulator studies. One reason for the difference is that when a TOR is issued prior to conflict onset, drivers are given the time needed for their preparatory actions (e.g., placing hands on the wheel, deactivating AD) that is not needed when driving with ACC or in manual driving (baseline), before having to respond to the conflict scenario. Thus, at conflict onset the drivers in AD are as ready to act (hands on wheel, eyes forward) as the drivers in the baseline and can perform an avoidance manoeuvre similar as to the baseline drive. Overall, the present study shows that AD does not need to end up in a highly critical situation if the TOR is issued early enough. In fact, AD with an early TOR may be safer than driving with ACC, because in the former drivers are more likely to brake earlier in preparation for the conflict. Finally, a TOR clearly communicates the need for drivers to resume manual control and handle potential events when AD has been deactivated. In our study, once the drivers had taken control, they clearly understood their responsibilities to respond to the conflict, in contrast to a previous study with a similar, near-perfect assisted driving system.     

How does emotional intelligence predict driving behaviors among non-commercial drivers?

Dangerous driving behaviors have been found to be a leading contributor to vehicle crashes and fatalities, with more than 2.7 million people injured and 36,560 people killed in the United States in 2018 (NHTSA, 2020). Drivers’ emotions have been found to be among the leading contributors to dangerous driving behaviors. Emotions can be measured and understood through one’s emotional intelligence (EI). Previous research has confirmed the relationship between EI and dangerous driving behaviors among general driving populations in limited scope. This study analyzed dangerous driving behaviors (e.g., aggressive driving) among non-commercial US drivers. 615 US drivers ages 18 to 65 (M = 31.14, SD = 11.15) with valid US driver’s licenses (non-commercial) participated in this study. Participants completed an online survey through Qualtrics that included the Trait Emotional Intelligence Questionnaire (TEIQue-SF) to measure different dimensions of EI and the Dula Dangerous Driving Index (DDDI) and the Driving Behavior Questionnaire (DBQ) to measure dangerous driving behaviors. Furthermore, participants reported their demographic information, including age, sex, and location. Correlation analysis revealed that significant associations exist between dangerous driving behaviors and EI. The emotionality component of EI was found to be the strongest predictor of dangerous driving behaviors. The findings concluded that participants with higher EI scores engaged in less dangerous driving behavior, resulting in fewer crashes and fatalities. Thus, promoting and improving EI may be useful in preventing risky driving among non-commercial drivers. Incorporating emotional intelligence education in driver’s education, workplace training, and licensing procedures can be helpful to develop safer drivers. Further research is needed to investigate commercial drivers’ behaviors in relation to EI.     

Please stop now,automated vehicle! – Passengers aim to avoid risk experiences in interactions with a crossing vulnerable road user at an urban junction

The success of highly automated vehicles (HAVs; SAE Level 4) will depend to a large extent on how well they are accepted by their future passengers. This is especially true for the interaction of these vehicles with other human road users in mixed traffic. In future urban traffic, passengers of HAVs will observe from a passive position how the automated system resolves space-sharing conflicts with crossing vulnerable road users (VRUs; e.g., pedestrians and cyclists) at junctions. For one such crossing-paths conflict, we investigated when passengers would want the HAV to start braking and how much perceived risk passengers accept in the interaction of their vehicle with VRUs. To this end, we conducted 1) an online video study (N = 118), 2) a driving simulator study (N = 28), and 3) a human&vehicle-in-the-loop (Hu&ViL) study at a test site (N = 10). We varied the speed of the HAV (30 km/h vs. 50 km/h), the type (cyclist vs. pedestrian), and crossing direction of the VRU (left vs. right). During the approach to the junction, passengers' task was to trigger the HAV's braking maneuver, in a first trial at the point they considered ideal and in a second trial at the last point they still considered safe enough to decelerate and come to a stop at the stop line. For each braking maneuver, we analyzed the HAV’s distance and time-to-arrival (TTA) to the VRU at braking onset, as well as passengers’ perceived risk in the VRU interaction. The results showed that most passengers preferred harmless interactions with VRUs (at the ideal braking onset time), and accepted unpleasant, but not dangerous interactions at most (at the last acceptable braking onset time). Methodologically, the results were very similar in the three different environments (online, driving simulator, real vehicle). These results clearly show that, in addition to the technical considerations of safe automated driving, passengers’ perception and evaluation of HAV driving behavior should also be taken into account to achieve a satisfying level of acceptance of these vehicles.