Advanced driver assistance systems (ADAS) are vehicle technology features that are designed to increase the safety by providing warnings, e.g., forward collision warning (FCW), and, in some cases, intervening to avoid an unsafe situation (e.g., automated emergency braking (AEB) and lane keeping assist (LKA). These features are becoming more and more prevalent. NHTSA announced that ten manufacturers installed AEB on all new vehicles from fall 2019 through summer 2020 (NHTSA, 2020). Additionally, more than 90% of all new vehicles are equipped with adaptive cruise control (ACC) and half are able to provide both steering and speed control (Bartlett, 2020).

There is concern that ADAS users may misunderstand or overestimate a system’s capabilities, or incorrectly believe ADAS features to be technologies that enable vehicles to be autonomous or self-driving. Research has shown that some owners of vehicles with ADAS features don’t understand what systems their vehicle has (Harms et al., 2020), what their purpose or limitations are (e.g., McDonald, et al., 2018; DeGuzman and Donmez, 2021), and they may choose to disable some of the ADAS features on their vehicles (Reagan & McCartt, 2016). How consumers learn about ADAS technologies can have an impact on their understanding and perceptions of the features (Reyes et al., 2017; Nylen et al., 2019). Branding of ADAS features with system information that emphasized driver convenience rather than driver responsibility led to overconfidence in the system’s capabilities (Singer & Jenness, 2020). A high level of understanding and knowledge for ACC resulted in better driver performance in a recent simulator study (Gaspar et al., 2020). An incorrect mental model for or an inappropriate amount of trust in an ADAS feature may contribute to crashes while appropriate models may lead to prevented or mitigated crashes.

While some studies have used police-reported crashes to evaluate the effectiveness of ADAS technologies (Fildes et al., 2015; Cicchino, 2018), there are limitations associated with obtaining the data in this manner. Contrary to conventional understanding, having the vehicle make, model and vehicle identification number (VIN) does not ensure that the technologies available on the vehicle can be identified. Many times, these systems are an option or part of a package that may or may not have been purchased. An officer investigating a crash may interview the driver about what technologies are present on the vehicle and whether they were in use, but this approach can also be problematic if either the driver or the officer is unfamiliar with or doesn’t clearly understand the ADAS features. Finally, even when an investigating officer considers whether ADAS are present on the vehicle and may have played a role in the crash, the crash reporting structure does not readily facilitate the inclusion of that information. As a result, there is a dearth of information about the circumstances of crashes involving vehicles with ADAS features and the mental models of the motorists who were operating those vehicles.

This research aims to help fill the gaps in knowledge about the impact of ADAS features in real-world crashes occurring in Iowa and Colorado, including whether the driver had an accurate mental model of the vehicle’s ADAS feature(s), crash characteristics such as the environmental and traffic conditions, and whether the feature(s) potentially contributed to or mitigated the crash. Interview guides to facilitate the gathering of information from motorists operating vehicles likely to be equipped with ADAS features that have been involved in crashes will be developed. Another guide will be developed in order to interview officers who have investigated crashes of ADAS-equipped vehicles. Qualitative analysis will be conducted and a final report will present findings and recommendations for various stakeholders, including infrastructure owners/operators, law enforcement, crash reporting agencies, departments of transportation, and original equipment manufacturers (OEMs).

List of tasks include:

1. IRB application, TAC formation, and data use agreements
2. Develop the motorist interview guide
3. Develop the investigating officer interview guide
4. Identify and invite potential motorist respondents
5. Interview motorists
6. Interview investigating officers
7. Conduct analysis

Colorado and Iowa DOTs will each contribute $75,000 to initiate the project.