A WOZ Study of Feedforward Information on an Ambient Display in Autonomous Cars Hauke Sandhaus Bauhaus-Universität Weimar Fac. of Media, Human-Computer-Interaction 99423 Weimar, Germany hauke.sandhaus@uni-weimar.de Eva Hornecker Bauhaus-Universität Weimar Fac. of Media, Human-Computer-Interaction 99423 Weimar, Germany eva.hornecker@uni-weimar.de ABSTRACT We describe development and user testing of an ambient dis- play for fully autonomous vehicles. Instead of providing feed- back about driving actions once executed, it communicates driving decisions in advance, via light signals in passengers’ peripheral vision. This was tested in an WoZ-based on-the- road-driving emulation of a autonomous vehicle. Findings from a preliminary study with 14 participants suggest that such a display might be particularly useful to communicate upcoming inertia changes for passengers. CCS Concepts •Human-centered computing → Human computer inter- action (HCI); Interface design prototyping; HCI design and evaluation methods; Author Keywords Autonomous Vehicle Interfaces; Methodology; On-Road simulation; Ambient Display; INTRODUCTION HCI research on passengers in fully autonomous vehicles is sparse, and reasons are multifold: access to real autonomous vehicles is limited as few companies have working prototypes. Current research focuses on semi-autonomous cars, as these will be sooner available and have urgent HCI issues to be solved (such as hand-over scenarios or collaborative driving [15, 13]). Finally, fully autonomous vehicles appear to have few user interactions beside selecting the pick-up and drop-off location. Some work [16, 1] addressed accessibility to self- driving vehicles with the RRADS driving platform: our study presented here used a modified version which seats passen- gers in the back seat to simulate fully autonomous driving. Being a passenger in a car or bus is arguably not as enjoy- able as traveling on a modern train. One reason might be the acceleration-deceleration behavior [2], and self-driving vehi- cles might worsen this because of frequent inertia changes due Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). UIST ’18 Adjunct October 14–17, 2018, Berlin, Germany © 2018 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-5949-8/18/10 DOI: https://doi.org/10.1145/3266037.3266111 Control -Tablet Control -Unit Interaction Wizard Driver Subject -Tab Instruction & Questionaire App Curtain Figure 1. The study setup to falsely detected obstacles[11]. Furthermore, many passen- gers exhibit back-seat driver behavior[3], and we do not know how they will behave faced with total loss of control. The cabin of an autonomous cab should be practical and com- fortable. Passengers might adjust their seats for comfort (lying down, seated backward, facing each other or turned away). They might prefer to use their own devices over the cars’ info- tainment system. Therefore, unnecessary graphical interfaces should be avoided [8]. Instead of putting screens everywhere, our proposed ambient display turns the cabin itself into a sur- face for feedback, utilizing colored light to convey information. Ambient displays have been used in vehicles to keep drivers’ eyes on the road [9, 12, 10]. ’Feedforward’ is usually used in context with affordances of interfaces, here it is used as ’advance feedback before executing an action’ [7, 4, 14]. OUR PROTOTYPE Our ambient display augments the driving decisions of the vehicle with light animations before they are carried out. It visualizes both planned events such as taking turns, speeding up and slowing down, as well as unplanned events such as swerving because of unexpected obstacles. The events are ani- mated smoothly and mapped to the shape of the light display: urgent events use stronger light, colors and rapid animations whereas less important events are animated subtle. Where