Assessing Drivers’ Trust of Automated Vehicle Driving Styles With a Two-Part Mixed Model of Intervention Tendency and Magnitude

Abstract

Objective: This study examines how driving styles of fully automated vehicles affect drivers’ trust using a statistical technique—the two-part mixed model—that considers the frequency and magnitude of drivers’ interventions. Background: Adoption of fully automated vehicles depends on how people accept and trust them, and the vehicle’s driving style might have an important influence. Method: A driving simulator experiment exposed participants to a fully automated vehicle with three driving styles (aggressive, moderate, and conservative) across four intersection types (with and without a stop sign and with and without crossing path traffic). Drivers indicated their dissatisfaction with the automation by depressing the brake or accelerator pedals. A two-part mixed model examined how automation style, intersection type, and the distance between the automation’s driving style and the person’s driving style affected the frequency and magnitude of their pedal depression. Results: The conservative automated driving style increased the frequency and magnitude of accelerator pedal inputs; conversely, the aggressive style increased the frequency and magnitude of brake pedal inputs. The two-part mixed model showed a similar pattern for the factors influencing driver response, but the distance between driving styles affected how often the brake pedal was pressed, but it had little effect on how much it was pressed. Conclusion: Eliciting brake and accelerator pedal responses provides a temporally precise indicator of drivers’ trust of automated driving styles, and the two-part model considers both the discrete and continuous characteristics of this indicator. Application: We offer a measure and method for assessing driving styles.

Effect of Environmental Factors and Individual Differences on Subjective Evaluation of Human-like and Conventional Automated Vehicle Controllers

“…Currently, the main focus of most vehicle manufacturers is on addressing the technical challenges required for developing safe and reliable automated vehicles for deployment on the road. However, although safety is one of the most important aspects of vehicle design, ensuring that future automated vehicles are used, and accepted, by consumers also relies on creating a comfortable ride experience for the on-board user (Bellem, Thiel, Schrauf, & Krems, 2018;Lee, Liu, Domeyer, & DinparastDjadid, 2019;Voß, Keck, & Schwalm, 2018). The challenge with automated vehicles that do not consider the users' experience and personal preferences in the design process is not just that they create an unpleasant experience for the occupant, but that they may affect the trust, acceptance, and consistent use of the system, by both drivers, and other road users (Price, Venkatraman, Gibson, Lee, & Mutlu, 2016;Siebert, Oehl, Höger, & Pfister, 2013).…”

Designing for the Extremes: Modeling Drivers’ Response Time to Take Back Control From Automation Using Bayesian Quantile Regression

“…However, future studies could apply more complex models, such as a mixture model, to address this limitation. Nonresponses after a TOR could also be added to the analysis of the continuous response distribution using models such as the two-part model that is used to study discrete-continuous data (Belotti et al, 2015; Lee et al, 2019).…”

What drives citations of frontier application publications?

“…Handling such data using linear regression without taking cognisance of the many zeros could pose challenges regarding the difficulties in deriving error terms that are normal ( Belotti et al., 2015 ; Carboni, 2012 ). We used two approaches in the literature to address the problem: inverse hyperbolic sine transformation of the dependent variable ( Bellemare and Wichman, 2019 ; Burbidge et al., 1988 ; Lavin et al., 2017 ; MacKinnon and Magee, 1990 ; Pence, 2006 ) and two-part mixture models ( Cragg, 1971 ; Ciminata et al., 2020 ; Duan et al., 1984 ; Lee et al., 2019 ; Mihaylova et al., 2011 ).…”

“…Subsequently, based on a positive outcome, a regression model is estimated for the positive outcome ( Belotti et al., 2015 ; Ciminata et al., 2020 ; Zhao et al., 2020 ). Following Lee et al. (2019) , let …”