Address inputting while driving: a comparison of four alternative text input methods on in-vehicle navigation displays usability and driver distraction.

OBJECTIVE

Efficient and safe address entry is crucial to in-vehicle navigation systems. Although various text input methods (TIMs) are commercially available, to date, the details of the driver's interactions with these TIMs in the vehicle are poorly understood. Therefore, the effect of four alternative TIMs conditions on in-vehicle navigation displays usability and driver distraction were directly compared. For reference, the baseline driving task (distraction-free) was also investigated.

METHODS

A city expressway simulator experiment including 25 young drivers was launched. Under each condition, the driving task was lane-keeping with speed ranging between 40 and 60 km/h, and the navigation task was to enter a 14-characters Chinese address name. In the meantime, usability (text entry time, number of errors, and preference) and driver distraction (NASA-TLX, average speed, the standard deviation of lane position, total glance duration, number of glances, average glance duration, and number of glances exceeding 1.6 s) metrics were measured as dependent variables. A sequence of one-way repeated measure analyses of variance (ANOVA) was performed to examine which type of TIMs can maximize in-vehicle navigation displays usability and minimize driver distraction.

RESULTS

Generally, lateral driving performance deteriorated with the addition of the address inputting task, and the four alternative TIMs might fall into three levels: Speech is optimal, Qwerty followed, Shape-writing and Handwriting ranked last. Specifically, word-based speech remains performed best on all observed metrics for Chinese address names. There was an insignificant difference in text entry time and total glance duration among Qwerty, Shape-writing, and Handwriting. However, Shape-writing and Handwriting are not suitable for young drivers since the nature of uninterruptible causes excessive errors, more considerable lane position variation, longer average glance duration, and more glances exceeding 1.6 s.

CONCLUSIONS

This study provides valuable insights into young drivers' interactions with four alternative TIMs. Importantly, it is beneficial to the automotive user interfaces design of in-vehicle navigation displays and other sub-functions of in-vehicle information systems (IVISs), such as music playback and text messaging, which positively mitigate driver distraction and prevent traffic injuries.