Drivers of partially automated vehicles are blamed for crashes that they cannot reasonably avoid

Abstract

People seem to hold the human driver to be primarily responsible when their partially automated vehicle crashes, yet is this reasonable? While the driver is often required to immediately take over from the automation when it fails, placing such high expectations on the driver to remain vigilant in partially automated driving is unreasonable. Drivers show difficulties in taking over control when needed immediately, potentially resulting in dangerous situations. From a normative perspective, it would be reasonable to consider the impact of automation on the driver's ability to take over control when attributing responsibility for a crash. We, therefore, analyzed whether the public indeed considers driver ability when attributing responsibility to the driver, the vehicle, and its manufacturer. Participants blamed the driver primarily, even though they recognized the driver's decreased ability to avoid the crash. These results portend undesirable situations in which users of partially driving automation are the ones held responsible, which may be unreasonable due to the detrimental impact of driving automation on human drivers. Lastly, the outcome signals that public awareness of such human-factors issues with automated driving should be improved.

Figure 1

Two example vignette visuals of (left) an intentionally distracted driver engaging with the vehicle’s entertainment center and (right) an unintentionally distracted driver whose mind is wandering. See the Supplementary methods for all vignettes.

Figure 2

Responsibility attributed to each actor by the participants for all factor levels (distraction and cause of distraction). Data are visualized using violin plots, box plots, and individual data points. Cause of distraction was only varied within the distracted factor levels.

Figure 3

Driver’s situation awareness and ability to take control as perceived by the participants per distraction level and source of distraction. Data are visualized using box and violin plots.

Figure 4

Median responsibility attribution to the driver and the manufacturer per code identified in the thematic analysis of the participants’ reasoning. Codes that were mentioned at least 10 times are visualized here for clarity (see Supplementary figures 17 and 18 for the other codes). The number of times the argument was made is included in brackets. The lines indicate 95% confidence interval of the median responsibility attributions to the driver and manufacturer for each code.

Figure 5

Attributed responsibility versus driver ability to take control and avoid the crash. The distribution of the responses for driver ability and corresponding attributed responsibility per participant are visualized using a kernel density estimate plot (Gaussian kernels, contour threshold at 0.25; e.g. 75% of the probability mass is indicated in the shaded areas) for the intentional and unintentional factor levels (short and long distraction factor levels pooled). The shaded areas represent 75% of the data probability mass per group. The black identity line is a qualitative representation of the normative expected attribution of responsibility given the driver’s ability to take control; attributed responsibility should be equal or lower to the driver’s control ability.

Figure 6

The conceptual model; the corresponding statistical model is shown in Supplementary Fig. 1.