Home // ACHI 2020, The Thirteenth International Conference on Advances in Computer-Human Interactions // View article

Trust Me! I can be a Designated Driving Assistant

Authors:
Misbah Javaid
Vladimir Estivill-Castro
Rene Hexel

Keywords: Human-Robot Trust; Explanations; Queensland Traffic Rules; Human-Robot Physical Interaction.

Abstract:
Autonomous Vehicles drive themselves by utilizing sensors and artificial intelligence. Evidence from surveys have shown that people are captivated by AVs, yet reluctant to give up control entirely to the Autonomous Vehicles. Inadequacy of trust has been identified as a pre-eminent factor behind this unacceptability. We propose that explanations describing behavioural decisions serve to upgrade human’s sense of trust on the driving performance of Autonomous Vehicles. The contribution of our proposed research is tested by creating a human-robot interactive scenario with 34 human participants, in which we present a robot as a Driver of an Autonomous Vehicle. The Robot Driver is an expert in recognising and explaining traffic regulations and traffic signs. Moreover, the Robot Driver is equipped with the ability to make decisions on uncertain road problem-solving situations in terms of explaining, i.e., what is the best decision and why; keeping in view traffic regulations. Additionally, the Robot Driver is competent judging when to overstep a traffic regulation, relative to a perceived hazard on the road. During the interactive scenario, human participants solve different road problem-solving scenarios with the help of the Robot Driver. We examined the effect of explanations from the Robot Driver on human’s trust under two conditions (1) no-error (2) error-justification and correction. Overall, the results indicate that during the problem-solving task, the human participants trusted and conformed more with the Robot Driver’s decisions as compared to their own decisions. Furthermore, the human participants perceived the behaviour of the Robot Driver under Condition 1 as very reliable, intelligent and trustworthy for driving than under Condition 2. We conclude explanations are an effective communication modality that not only improved the perceived agency of the Robot Driver but also improved trust of the human participants on the Robot for driving. Therefore, explanations that disseminate behavioural decisions of AVs can help in improving human’s trust in them.

Pages: 164 to 171

Copyright: Copyright (c) IARIA, 2020

Publication date: March 22, 2020

Published in: conference

ISSN: 2308-4138

ISBN: 978-1-61208-761-0

Location: Valencia, Spain

Dates: from November 21, 2020 to November 25, 2020