Evaluating feedback requirements for trust calibration in automated vehicles

Abstract

The inappropriate use of automation as a result of trust issues is a major barrier for a broad market penetration of automated vehicles. Studies so far have shown that providing information about the vehicle’s actions and intentions can be used to calibrate trust and promote user acceptance. However, how such feedback could be designed optimally is still an open question. This article presents the results of two user studies. In the first study, we investigated subjective trust and user experience of (N=21) participants driving in a fully automated vehicle, which interacts with other traffic participants in virtual reality. The analysis of questionnaires and semi-structured interviews shows that participants request feedback about the vehicle’s status and intentions and prefer visual feedback over other modalities. Consequently, we conducted a second study to derive concrete requirements for future feedback systems. We showed (N=56) participants various videos of an automated vehicle from the ego perspective and asked them to select elements in the environment they want feedback about so that they would feel safe, trust the vehicle, and understand its actions. The results confirm a correlation between subjective user trust and feedback needs and highlight essential requirements for automatic feedback generation. The results of both experiments provide a scientific basis for designing more adaptive and personalized in-vehicle interfaces for automated driving.

Investigating explanations in conditional and highly automated driving: The effects of situation awareness and modality

“…The AV, as a complex AI system, also needs to be explained for better human-AV team performance, since it is important to keep an appropriate level of trust in automation and effectively manage uncertainty. Previous studies already confirmed the necessity of feedback in autonomous driving (Seppelt & Lee, 2019;Wiegand et al, 2020;Wintersberger, Janotta, Peintner, Löcken, & Riener, 2021). For example, Wintersberger et al (2021) found that regardless of the trust in AV, people still preferred to be informed about forthcoming strategies and maneuvers.…”

“…Previous studies already confirmed the necessity of feedback in autonomous driving (Seppelt & Lee, 2019;Wiegand et al, 2020;Wintersberger, Janotta, Peintner, Löcken, & Riener, 2021). For example, Wintersberger et al (2021) found that regardless of the trust in AV, people still preferred to be informed about forthcoming strategies and maneuvers.…”

Is the automated vehicle “aware” of the pedestrian? Examining driving behavior adaptation as a cue to inform the passenger of a potential hazard

“…Previous research has examined various aspects and underlying determinants of passenger trust in automated vehicles, including the requirements and effects of feedback from the automated driving system via in-vehicle human-machine interfaces (HMIs; e.g., Colley et al, 2020;Feierle et al, 2020;Hartwich et al, 2021;Wintersberger et al, 2021;Yun & Yang, 2022;see Ekman et al, 2017 for a trust-based framework for HMI design), uncertainty displays (e.g., Beller et al, 2013;Helldin et al, 2013), automation malfunctions (e.g., Feierle et al, 2021), overtrust (e.g., Victor et al, 2018), driving behavior and style (e.g., Ekman et al, 2019;Mühl et al, 2020), and initial or introductory information (e.g., Forster et al, 2018;Körber et al, 2018) on self-reported passenger trust.…”

A Forgotten Danger in DNN Supervision Testing: Generating and Detecting True Ambiguity

“…The existence of uncertainty led to the development of DNN Supervisors (in short, supervisors), which aim to recognize inputs for which the DL component is likely to make incorrect predictions, allowing the DLS to take appropriate countermeasures to prevent harmful system misbehavior [2]- [8]. For instance, the supervisor of a self-driving car might safely disengage the auto-pilot when detecting a high uncertainty driving scene [2], [9]. Other examples of application domains where supervision is crucial include medical diagnosis [10], [11] and natural hazard risk assessment [12].…”