Promoting sustainable mobility requires technological innovation and changes in individual travel behavior. Using the Theory of Planned Behavior, we examined how attitudes, norms, and perceived control shape the willingness to adopt alternatives to car use. We designed 38 future commuting scenarios, each of which isolated a single dimension across three mobility concepts: public transportation, cycling, and shared automated vehicles. In an online survey (N = 168), participants rated their willingness to switch modes and pay more. To deepen our understanding, we conducted follow-up interviews (N = 10), exploring their everyday mobility practices and their likes and dislikes regarding the practicality of the future scenarios. Our findings show that features linked to instrumental attitudes and control beliefs elicit stronger intentions than affective cues, ecological appeals were less persuasive. We argue that effective behavior change depends on linking motivational factors to the realities of everyday mobility contexts.

The Development of Fully Autonomous Vehicles (AVs) would fundamentally change the nature of in-vehicle user interactions, behaviors, needs, and activities. Passengers free from driving would expect to undertake diverse Non-Driving-Related Tasks to keep themselves occupied. Introducing Conversational Artificial Intelligence (CAI) in Level 5 AVs could improve the in-vehicle user experience (UX). To explore this, firstly, we identify what roles and relationships can CAI play towards end-users of AVs through end-user interviews and thematic analysis. Secondly, we examine how end-users qualitatively assess the embodied UX of the CAI roles and relationships through guided brainstorming, post simulator interaction experiments employing Wizard of Oz setup and Participant Enactment methods. Results show that Tour Guide, Mentor, and Storyteller were the most preferred CAI roles, and that Human-CAI relationships are maintained if the CAI mediates in-vehicle user activities, interactions, sharing of vehicle control, and deep conversations. We discuss the research implications and propose design guidelines.