Mediated participatory design for contextually aware in-vehicle user-experiences with autonomous vehicles

Abstract

This study reports on the empirical findings of a series of participatory design workshops for the development of a supportive automotive user experience design system. Identifying and addressing this area with traditional research methods is problematic due to the different user experience (UX) design perspectives that might conflict and the related limitations of the automotive domain. Consequently, we deploy a pragmatic epistemological paradigm and apply participatory prototyping methods to resolve this problem. We conduct two iterations of design and evaluation with 19 user experience (UX) designers through individual participatory prototyping activities to gain insights into their explicit, observable, tacit and latent needs. We describe the design of a toolkit tailored to the character of the study to be used in relevant studies of ill-defined or wicked problems. The participatory design activities initially allowed us to explore the motivation to use different technologies, the system’s architecture, detailed features of interactivity, and to describe our users’ needs. As a result, our first analysis of data led us to design implications that translate participants’ needs into UX goals. We use these UX goals for the design of goal-directed personas and scenarios of use as actionable insights to develop our system. A medium-fidelity functional prototype of our system was then evaluated, while contextually aware automotive UX practitioners criticised our design decisions. Some of the essential findings when supporting the contextual understanding are generating new knowledge to inform both theory and practice. The results propose that most automotive UX designers are ready to adopt technologies that use sensitive physiological measures such as eyes, face, body tracking using cameras and computer vision. In contrast, non-automotive UX designers who empathise with the passengers and the drivers and perceive the in-vehicle space as something more private are suggesting that this might affect people’s trust. The majority agrees to collect data and communicate with the users using implicit and explicit context, as a way to support UX design in the autonomous vehicles would require the consent of the passengers. Even though UX designers suggested a general interest in the social and temporal context of the interactions, the limitations of privacy and safety in the vehicle limit them in collecting task-related contextual data leaving the social, temporal, and physical context unexplored. Safety is arguably a factor that will not restrict the future of autonomous driving experiences research and design since there is no cognitive demand on level five autonomy which hands the passengers with plenty of other options when not driving, assuming that they are ready to trust a fully automated system. However, our study does not provide us with a direction on the privacy of autonomous vehicle experiences and whether privacy will continue being a limitation in the context of self-driving vehicles. Thus, we would recommend further research on trust and privacy in fully automated vehicles. We conclude by discussing the design implications and functional tools of our system, including 1) a video tagging tool that supports saving an occurrence identified momentarily on real-time video. 2) A privacy call-wall which uses implicit and explicit context to avoid intrusiveness in private situations. 3) A human-like avatar tool for mitigating privacy issues, and 4) an interactive interviewing tool to support communication between UXers and the passengers of autonomous vehicles. Finally, 5) exploration tools, including a tool for searching participants’ characteristics and target groups of people. We further inform the body of knowledge in participatory UX and HCI methods about the advantages of our methodological approach and the limitations of using it. We discuss why involving non-experts in co-design activities using toolkits tailored to the domain of interest is valuable. Furthermore, we extensively address how, and we give directions for the design of similar toolkits by describing the toolkit that we designed and applied in our study. Conclusively we discuss the broader implications of trust and privacy in other domains and how this related to our findings.