Impact of User Interface Design on Driver Distractibility

Last year, there were over 40,000 vehicle deaths in the United States. In 2016, more than 9 percent of United States traffic deaths — or 3,450 — were linked to distracted driving.  In the last two decades, vehicles have continued pushing design and ergonomic boundaries with their controls, in pursuit of attractive, convenient, and safe interiors. Our test vehicles reflect these recent evolutions, with integrated, in- or on-dash multimedia systems and climate controls unlimited by design constraints of decades past. Despite their similarities and differences, these vehicles represent each respective automaker’s best efforts to improve ease-of-use and reduce distractibility from vehicle operation.

Our project used eye-tracking glasses to compare distractibility and usability of vehicle interfaces. We had five participants participate in 18 tasks across two vehicles: a 2005 Lexus LS and 2023 Honda Civic. 

We used a Pupil Invisible eye tracker for this test. The purpose of using the tracker was to explore the learnability and efficiency of a vehicle interface from 2005 compared to one from 2023. The tracker was also very helpful in functioning as a recording device generally - we got a first-person view of how the user perceived the interface. We hoped to measure the scanpath, number of glances (fixations), and length of time to complete tasks to identify biosensory factors that could clue us into vehicle UX design. 

Our five participants were between the ages of 27-32 and were required to have a driver’s license and experience driving a car within at least the last few weeks. We measured where participants were looking as they performed tasks related to HVAC (heating, ventilation, and air conditioning), audio, and navigation. We reviewed guidelines from The National Highway Traffic Safety Administration (NHTSA) for automakers that measure distractibility by “number of eye glances.” Depending on the task, the guidelines suggest 1.5-2 seconds of “visual-manual attention” and 9-12 seconds of cumulative time on a task (such as tuning a radio station or entering an address in a navigation system). We ended up with a list of 18 tasks and created a test script for each task. For all of these tasks, drivers were stationary - no actual driving took place. After tasks were completed in each car, we asked users a set of qualitative questions about their experience. The questions were designed to provide a qualitative lens to our study. 

Overall, we find that it takes significantly less time to complete the HVAC and navigation tasks in the 2023 Honda compared to the 2005 Lexus. Participants described the Honda to be more “intuitive” and “familiar” than the Lexus, suggesting that control type and placement allowed them to complete tasks more easily. However, we also find that it takes slightly longer to complete the audio tasks in the Honda compared to the Lexus. Possible reasons for this observation were not clearly discussed in participants’ post-test responses, but we attribute this to the Honda’s higher degree of audio choice and settings, as well as lack of dedicated, physical audio controls. 

Qualitative data reveal four of the five total participants found the Honda to be less distracting than the Lexus, supporting the quantitative data showing lower task completion times in the Honda. Although qualitative results show some participants taking much longer than others to complete tasks in both cars, participants mostly spoke of difficulties they encountered in the Lexus. User interfaces with “flat” navigational hierarchy, as well as tactile physical controls, appear more learnable to new users than those without.