Habits, attitudes, and expectations of regular users of partial driving automation systems

Convenience or safety system? Crash rates of vehicles equipped with partial driving automation

OBJECTIVE

Although partial driving automation systems are usually discussed as convenience features, consumers sometimes consider them to be safety features. The goal of this study was to assess if partial driving automation reduces rear-end and lane departure crashes beyond safety systems like automatic emergency braking (AEB) and lane departure prevention (LDP).

METHODS

Analyses examined crash rates of model year 2017-2019 Nissan Rogues and model year 2013-2017 BMW vehicles. Negative binomial regression was used to assess the association of Nissan's partial driving automation system, ProPILOT Assist, and BMW's system, Driving Assistant Plus, with police-reported rear-end and lane departure crash rates on the limited-access roads where they are designed to be used per vehicle mile traveled. Crash rates were also examined on roads with speed limits of ≤ 35 mph, where the systems were expected to have limited functionality and not be used much.

RESULTS

Equipment with BMW's Driving Assistant Plus was not associated with significantly lower crash rates than equipment with LDP alone. Rear-end crash rates were 26% lower on limited-access roads and 43% lower on roads with speed limits ≤ 35 mph for Nissan Rogues with ProPILOT Assist than for those with AEB alone. Similarly, lane departure crash rates were 25% lower for Nissan Rogues with ProPILOT Assist compared with those with LDP alone on limited-access roads, but were 31% lower on roads with speed limits ≤ 35 mph and 43% lower on limited-access roads in the dark. This brings into question if the lower crash rates associated with ProPILOT Assist can be attributed to use of the system, given that it would be activated infrequently on residential roads and that vehicles with it generally had better headlights than those unequipped.

DISCUSSION

There is no convincing evidence that partial driving automation is a safety system that is preventing crashes in the real world. Research incorporating system use will be key to understanding safety effects. Considering that drivers have been documented misusing these systems, designing partial driving automation with robust safeguards to deter misuse will be crucial to minimizing the possibility that the systems will inadvertently increase crash risk.

A conceptual framework for automation disengagements

A better understanding of automation disengagements can lead to improved safety and efficiency of automated systems. This study investigates the factors contributing to automation disengagements initiated by human operators and the automation itself by analyzing semi-structured interviews with 103 users of Tesla's Autopilot and FSD Beta. The factors leading to automation disengagements are represented by categories. In total, we identified five main categories, and thirty-five subcategories. The main categories include human operator states (5), human operator's perception of the automation (17), human operator's perception of other humans (3), the automation's perception of the human operator (3), and the automation incapability in the environment (7). Human operators disengaged the automation when they anticipated failure, observed unnatural or unwanted automation behavior (e.g., erratic steering, running red lights), or believed the automation is not capable to operate safely in certain environments (e.g., inclement weather, non-standard roads). Negative experiences of human operators, such as frustration, unsafe feelings, and distrust represent some of the adverse human operate states leading to automation disengagements initiated by human operators. The automation, in turn, monitored human operators and disengaged itself if it detected insufficient vigilance or speed rule violations by human operators. Moreover, human operators can be influenced by the reactions of passengers and other road users, leading them to disengage the automation if they sensed discomfort, anger, or embarrassment due to the automation's actions. The results of the analysis are synthesized into a conceptual framework for automation disengagements, borrowing ideas from the human factor's literature and control theory. This research offers insights into the factors contributing to automation disengagements, and highlights not only the concerns of human operators but also the social aspects of this phenomenon. The findings provide information on potential edge cases of automated vehicle technology, which may help to enhance the safety and efficiency of such systems.

Trust in automated vehicles: constructs, psychological processes, and assessment

There is a growing body of research on trust in driving automation systems. In this paper, we seek to clarify the way trust is conceptualized, calibrated and measured taking into account issues related to specific levels of driving automation. We find that: (1) experience plays a vital role in trust calibration; (2) experience should be measured not just in terms of distance traveled, but in terms of the range of situations encountered; (3) system malfunctions and recovery from such malfunctions is a fundamental part of this experience. We summarize our findings in a framework describing the dynamics of trust calibration. We observe that methods used to quantify trust often lack objectivity, reliability, and validity, and propose a set of recommendations for researchers seeking to select suitable trust measures for their studies. In conclusion, we argue that the safe deployment of current and future automated vehicles depends on drivers developing appropriate levels of trust. Given the potentially severe consequences of miscalibrated trust, it is essential that drivers incorporate the possibility of new and unexpected driving situations in their mental models of system capabilities. It is vitally important that we develop methods that contribute to this goal.