• We conducted a review of research on in-vehicle touchscreen safety and usability. • We identified 73 papers covering a range of research areas. • Touchscreens tend to perform worse than physical/voice control with some exceptions. • Research is needed on industry trends (larger screens, touchscreens for driving tasks) • Research on interactions between touchscreen design elements is also needed. In-vehicle touchscreens are becoming increasingly common, and in some cases replacing physical controls. Given their pervasiveness, minimizing the negative impacts of touchscreens on safety and usability is essential. Currently, there are no existing guidelines for the design of in-vehicle touchscreens. We conducted a rapid review of research investigating the safety and usability of in-vehicle touchscreens, the results of which can be used to inform such guidelines. We conducted a search in four databases (ACM, Scopus, SAE Mobilus, Transport Research International Documentation). Our focus was on driver touchscreen use in manual vehicles or those with driving assistance (i.e., up to SAE Level 2 driving automation). Studies were included if they investigated the effect of touchscreen design or age on safety/usability using inferential statistics. Studies investigating brought-in or aftermarket touchscreens were excluded. We identified and summarized the findings of 73 studies that investigated a range of topics related to touchscreen interaction: comparison with physical or voice controls, screen size, use of driving automation, task type and difficulty, age, and other design considerations, such as feedback and button/text size. Overall, touchscreens tended to have negative effects on driving performance, visual attention, and secondary task performance compared with voice control and physical controls, although there were exceptions for some tasks. There was limited research using larger touchscreens, which are becoming increasingly common, as well as a lack of systematic research on interactions between different touchscreen elements (e.g., feedback, task type, screen size).
DeGuzman et al. (Sat,) studied this question.