Designing an Information System that Supports Decision Making for Passengers of Autonomous Vehicles

Abstract

Background As the technology related to self-driving cars develops, new perspectives on how people acquire information and formulate decisions are becoming more critical to design the newinfotainment system. This study aims to identify the information that passengers will require invehicles with the level 4 or 5 stages of self-driving that have not yet been commercialized. Also, thisstudy examines system designs that help passengers to make decisions through experiments.Methods This study investigates the types of information required by passengers of selfdrivingcars and user behavior changes in the interior space of automobiles. We conducted apreliminary survey to investigate the type of information required. Based on the result of the firststudy, we drew a scenario by deriving the structure and concept of the new infotainment system.The system was designed to provide information through windshields, touchscreen displays, andinteractive agents that explore the system-user interactions in autonomous vehicles through userobservation experiments.Results We identified through preliminary research that nearby restaurants are preferredfor passengers of level 4 or 5 self-driving cars. Looking at the preference of information provisionmethods by passenger group when there are one, two, and three passengers, we found that thepreference for interactive agents was lowered in driving situations with others. In groups of three,the interactive agent is less frequently answered to preemptive utterances, but they tend to orderagents after having their own conversation. Windshield displays, which allow passengers to shareand communicate information together, have also been evaluated to help with quick choices anddecisions.Conclusions The preference for preemptive utterance or recommendation of interactive agentsvaries depending on the passenger group type. Also, passengers prefer a windshield display typethat allows them to make decisions while looking at information. The results suggest that differentinteractions should be applied when designing future infotainment systems