Search

Studies show that drivers’ intention to use automated vehicles is strongly modulated by trust. It follows that their benefits are unlikely to be achieved if users do not trust them. To date, most studies of trust in automated vehicles have relied on self-reports. However, questionnaires cannot capture real-time changes in drivers’ trust, and are hard to use in applied settings. In previous work, we found evidence that gaze behaviour could provide an effective measure of trust. In this study we tested whether combining gaze behaviour with Electrodermal Activity could provide a stronger metric. The results indicated a strong relationship between self-reported trust, monitoring behaviour and Electrodermal Activity: The higher participants’ self-reported trust, the less they monitored the road, the more attention they paid to a non-driving related secondary task, and the lower their Electrodermal Activity. We also found evidence that combined measures of gaze behaviour and Electrodermal Activity predict self-reported trust better than either of these measures on its own. These findings suggest that such combined measures have the potential to provide a reliable and objective real-time indicator of driver trust. © 2019 Elsevier Ltd

In Level 3 automated vehicles, drivers must take back control when prompted by a Take Over Request (TOR). However, there is currently no consensus on the safest way to achieve this. Research has shown that participants interact faster with an avatar when this “glows” in synchrony with participant physiology (heartbeat). We hypothesized that a similar form of synchronization might allow drivers to react faster to a TOR. Using a driving simulator, we studied driver responses to a TOR when permanently visible ambient lighting was synchronized with participants’ breathing. Experimental participants responded to the TOR faster than controls. There were no significant effects on self-reported trust or physiological arousal, and none of the participants reported that they were aware of the manipulation. These findings suggest that new ways of keeping the driver unconsciously “connected” to the vehicle could facilitate faster, and potentially safer, transfers of control.

Automated vehicles promise to improve road safety and increase travelling comfort. Yet, recent accidents have raised awareness of the challenges posed by the way humans interact with such systems. Many of these challenges present a common denominator: user trust. While a lack of trust may induce drivers not to use the automated vehicle’s functionalities, excessive trust can lead to dangerous outcomes, with drivers using the system in situations it cannot cope with. Successful and safe interaction between humans and automated driving systems requires trust calibration: the process of continuously aligning driver trust to the reliability of the automated driving system. To date, there have been few studies focussing on the way driver trust varies from situation to situation, and the way in which experience in a range of situations may lead to more appropriate trust levels. Studies of driver trust also face a number of methodological issues including a lack of on-road investigations, concerns regarding the validity of simulator-based research, and the lack of reliable real-time measurements for the assessment of user trust. This dissertation partially addresses these research gaps. Our findings show that, before on-road experience, drivers tend to overestimate the capabilities of vehicles equipped with Level 2 systems. After experiencing the vehicles in multiple scenarios, drivers had a better understanding of vehicles’ limitations, resulting in better calibrated trust. Following studies were performed in simulated environments. Through a first validation study, we made sure that our driving simulator elicited in our participants a strong “sense of presence” - the feeling of truly belonging in the virtual environment. A second study showed that users’ trust levels in specific situations did not consistently match evaluations of vehicle reliability given by engineers. Finally, results obtained through a third study showed that combining gaze behaviour with electrodermal activity provided an effective measure of trust in a simulated automated vehicle. The real question is not whether “to trust or not to trust” automated driving technology. What we should be asking is when an automated vehicle can be trusted, and what can be done to calibrate user trust with the actual capabilities of the system.

To maximize road safety, driver trust in an automated vehicle should be aligned with the vehicle’s technical reliability, avoiding under- and over-estimation of its capabilities. This is known as trust calibration. In the study reported here, we asked how far participant assessments of vehicle capabilities aligned with those of the engineers. This was done by asking the engineers to rate the reliability of the vehicle in a specific set of scenarios. We then carried out a driving simulator study using the same scenarios, and measured participant trust. The results suggest that user trust and engineer perceptions of vehicle reliability are often misaligned, with users sometimes under-trusting and sometimes over-trusting vehicle capabilities. On these bases, we formulated recommendations to mitigate under- and over-trust. Specific recommendations to improve trust calibration include the adoption of a more defensive driving style for first-time users, the visual representation of the objects detected by the automated driving system in its surroundings in the Human Machine Interface, and real-time feedback on the performance of the technology.

During Continuous Descent Approaches (CDAs) aircraft glide towards the runway resulting in reduced noise and fuel usage. Here, we investigated whether such landings cause less noise annoyance than a regular stepwise approach. Both landing types were compared in a controlled laboratory setting with a Virtual Community Noise Simulator (VCNS), using four audio samples: an overflight during a regular approach (2000 ft altitude) and three aircraft performing CDAs at respectively 3000, 4000 and 5000 ft. The samples at 2000 ft and 4000 ft were recorded at a countryside road, a 360° photo of which was used for the virtual visuals. The other two CDA samples were derived from the recording at 4000 ft. Participants were asked to rate all flyover samples twice while being immersed in the virtual environment. The CDA at 3000 ft was rated as most annoying, likely due to a longer overflight duration, followed by the regular descent and then the CDAs at 4000 and 5000 ft. As CDAs follow a fairly steady trajectory, it was estimated that they will increase annoyance within an area of approximately 2.5 km2, as compared to regular landings. Outside of this area, CDAs may instead result in less annoyance than regular landings. © 2017 Elsevier Ltd

In road-crossing situations involving negotiation with approaching vehicles, pedestrians need to take into account the behavior of the car before making a decision. To investigate the kind of information about the car that pedestrians seek, and the places where do they look for it, we conducted an eye-tracking study with 26 participants and analyzed the fixation behavior when interacting with a manually-driven vehicle that approached while slowing and displaying a yielding behavior. Results show that a clear pattern of gaze behavior exists for pedestrians in looking at a vehicle during road-crossing situations as a function of the vehicle’s distance. When the car is far away, pedestrians look at the environment or the road space ahead of the car. With the approach, the gaze gradually shifts to the windshield of the car. We conclude by discussing the implications of this insight in the user-centered-design of optimal external Human-Machine-Interfaces for automated vehicles.

Overtrust and undertrust are major issues with partially automated vehicles. Ideally, trust should be calibrated ensuring that drivers’ subjective feelings of safety match the objective reliability of the vehicle. In the present study, we examined if drivers’ trust toward Level 2 cars changed after on-road experience. Drivers’ self-reported trust was assessed three times: before having experience with these vehicles, immediately after driving two types of vehicles, and two weeks after the driving experience. Analysis of the results showed major changes in trust scores after the on-road driving experience. Before experiencing the vehicles, participants tended to overestimate the vehicle capabilities. Afterwards they had a better understanding of vehicles’ limitations, resulting in better calibrated trust.

A crucial factor influencing the validity of VR is sense of presence. The aim of this study is, firstly, to increase sense of presence of simulator driving by introducing an anxiety-inducing risk factor. Secondly, to add clarification to the partly inconclusive relationship between sense of presence, anxiety and physiological markers. The study consisted of two groups. The control group drove in the simulator normally. The threat group expected an electric shock in case of collision. Self-report ratings (on sense of presence, anxiety and task load), behavioral (speed, steering reversal rate, brake pedal use) and physiological measurements (EDA, HR) were taken. We expected an increase in the sense of presence, arousal, steering reversals, brake pedal use, task load and anxiety ratings, and a decrease of speed in the threat group. Analyses revealed partial group differences in speed and steering reversal rate. Sense of presence was identical and notably high in both groups. This implies that either a more aversive stimulus might be needed to induce anxiety and increase sense of presence, or that it can hardly be increased at all in highly immersive VR systems.