Calibration of Trust in Automated Driving: A Matter of Initial Level of Trust and Automated Driving Style?

Abstract

Objective Automated driving is becoming a reality, and such technology raises new concerns about human–machine interaction on road. This paper aims to investigate factors influencing trust calibration and evolution over time. Background Numerous studies showed trust was a determinant in automation use and misuse, particularly in the automated driving context. Method Sixty-one drivers participated in an experiment aiming to better understand the influence of initial level of trust (Trustful vs. Distrustful) on drivers’ behaviors and trust calibration during two sessions of simulated automated driving. The automated driving style was manipulated as positive (smooth) or negative (abrupt) to investigate human–machine early interactions. Trust was assessed over time through questionnaires. Drivers’ visual behaviors and take-over performances during an unplanned take-over request were also investigated. Results Results showed an increase of trust over time, for both Trustful and Distrustful drivers regardless the automated driving style. Trust was also found to fluctuate over time depending on the specific events handled by the automated vehicle. Take-over performances were not influenced by the initial level of trust nor automated driving style. Conclusion Trust in automated driving increases rapidly when drivers’ experience such a system. Initial level of trust seems to be crucial in further trust calibration and modulate the effect of automation performance. Long-term trust evolutions suggest that experience modify drivers’ mental model about automated driving systems. Application In the automated driving context, trust calibration is a decisive question to guide such systems’ proper utilization, and road safety.

It seems smart, but it acts stupid. Development of trust in AI advice in a repeated legal decision-making task

“…At the same time, some comparable studies show that trust grows over time: Chiou et al [15] found that individuals learn to trust an intelligent (robotic) system as operators understand how to work with it over the course of several interactions successfully. Similarly, Manchon et al [50] assessed trust in an automated driving system over time (three assessments in 4 months). In contrast to earlier studies [41], trust increased over time for both trustful and distrustful drivers.…”

“…Manchon and colleagues posit that several positive interactions in the early phase of the study supported even distrustful participants to gain trust fast. It can be agreed upon that initial trust is crucial for longer interactions, especially motivating individuals to start using automated systems in the first place [16,50]. Tolmeijer et al [67] add that initial trust is important to accept better later mistakes, strengthening long-term interactions.…”

Why Does the Automation Say One Thing but Does Something Else? Effect of the Feedback Consistency and the Timing of Error on Trust in Automated Driving

“…Here, the focus was set on dynamic learned trust to investigate how TiAD evolves when drivers actually interact with the HAD system. Previous experiments showed drivers' TiAD was negatively influenced by critical situations (i.e., a driving situation that requires the automation to perform an urgent maneuver in order to preserve passengers' safety) occurring during drives [12][13][14]. Hoff and Bashir [8] reported the timing of error, which is related to the automation performance, had a major influence on the operator's trust development.…”

“…Critical situations that occur early during the interaction with automation have a greater negative influence than errors occurring after a positive experience with the automation [15,16]. In the driving context, errors may encompass a variety of automation malfunctions, such as unexpected Take-Over Requests (TORs) that decrease TiAD [12] or poorly managed situations [13].…”

Measurement and Analysis of Dynamic Trust in Automated Vehicles Lessons Learned, Recommendations, and Pilot Results

“…Meta-analyses have highlighted that subsequent trust dynamics are influenced by indiviudal's level of fatigue, situation awareness, and workload (Schaefer et al, 2016). Specific to trust in AVs, surveys have shown that gender and age influence individuals' intent to adopt AVs (Hohenberger et al, 2016;Hulse et al, 2018;Kyriakidis et al, 2014), that initial trust can be influenced by training materials provided to drivers (Hohenberger et al, 2016), and that initial trust affects trust dynamics (Manchon et al, 2021). To account for these factors, we sampled across age, sex, and dispositional trust.…”