User Acceptance in Autonomous Vehicles: The Evolution of the End User

摘要

Initial investigation into trust of autonomous vehicles, shows somewhat of a polarized attitude between people who would currently accept and people who would thoroughly reject these systems. These results were surprising since neither group had any actual personal experience with an autonomous vehicle. This raised questions such as what factors initially influenced these attitudes? Can an actual experience have an impact on these factors? Can this interaction between experience, attitudes, and evolution be modeled?In order to begin to answer these questions two novel constructs were developed and are proposed in this work. First, the Safety-Critical Technology Acceptance Model (SCTAM) was developed utilizing acceptance models from information technology and provides a mechanism to understand, model, and ultimately quantify user attitude toward use of safety-critical vehicle systems. Second, the virtual reality (VR) based Electronic Car Learning and Intelligence Program Simulator (ECLIPS) was designed to give the experience of using an autonomous automobile to gauge the influence this experience would have on user acceptance.Study results produced in this work show a 22% increase in overall system acceptance after an autonomous vehicle demonstration was given using ECLIPS, as well as a 44% increase in the user reported problems/benefits relating to the use autonomous automobiles. These very promising results show that VR simulation can be used by developers to effect end user acceptance, even before actual vehicle implementation, and can provide data to be used to steer the development of autonomous vehicles in the direction towards optimal user acceptance upon vehicle rollout.