No abstract

Future development of in-vehicle voice assistants (VA) aims at more adaptive and interactive interaction with users. One focus area is the development of proactive behaviour of VA. With the changing role from driver to passenger in increasingly automated vehicles, new opportunities for interactions with assistants in the car arise. However, potential use cases for proactive VA in automated vehicles have not been investigated so far. We present our approach to brainstorm and prioritize use cases for a proactive VA based on a specific persona and user journey, involving experts and users. Secondly, we present user’s assessment and reflections on the prioritized use cases after experiencing them in a driving simulator study. The findings show that especially office-work related use cases that enable an efficient use of the ride time or offers contributing to user’s well-being receive high acceptance from users. The analysis of the qualitative feedback highlights that for a proactive VA adaptivity to the user’s behaviour and non-intrusive formulation of suggestions and questions is important. The study provides a starting point to investigate proactive behaviour of in-vehicle VA in more detail. The proposed approach for use case derivation can also be applied to other personas or different application domains of VA.

No abstract

Automated vehicles (AVs) are expected to enable users to engage in non-driving-related activities (NDRAs). However, users do not easily trust an automated vehicle which poses new challenges for automotive human-machine interfaces (HMIs). Over-presenting vehicle information can distract users from NDRAs, and under-presentation can impact trust and user experience (UX) negatively. To investigate how to best present vehicle information to foster users' trust and UX while performing NDRAs, we designed two in-vehicle HMI concepts: 1) A colored and animated light bar display around the windshield and 2) a windshield display interface presenting pictograms and numbers. Results from a simulator study ($N=18$) indicate that both concepts contribute to a high trust level and UX while not affecting the NDRA performance compared to the baseline of not showing vehicle information. In addition, the light bar provides better UX than the windshield display and is also preferred by users. With our findings, we contribute to the effective design of presenting vehicle information in automated vehicles.

Das FlexCAR besteht aus einer autonom gesteuerten Fahrzeugplattform für die Mobilität von morgen, welche Use-Case-gesteuert als Forschungsdemonstrator fungiert, um neue technologische Features unmittelbar aus dem Forschungsstadium nach dem Plug-and-Play–Prinzip gezielt zu implementieren. Damit kann eine frühzeitige Validierung im Hinblick auf ein künftiges Anwendungspotenzial ermöglicht werden. Offene Soft- und Hardwareschnittstellen sind hier berücksichtigt oder werden weiterentwickelt.   FlexCAR comprises an autonomously controlled vehicle platform for the mobility of tomorrow. Use case-guided as a research platform, it enables implementing new technological features according to the plug-and-play principle. It allows for an early validation concerning potential future applications. Open software and hardware interfaces are included and will be further developed.