Over recent years, there has been a lively debate about whether (automated) vehicles should be equipped with novel external human-machine interfaces (eHMIs) to support interactions with vulnerable road users. Explicit signals, such as a frontal brake light (FBL) that communicates vehicle deceleration to road users in front of the vehicle, have been shown to substantially speed up deceleration recognition. However, emerging evidence suggests that introducing such signals may also produce unintended side effects, particularly when road users must interpret vehicles without eHMIs. To examine how FBLs shape perceptual processing under partial deployment, we conducted a laboratory experiment combining response times and eye tracking. Twenty-nine participants viewed 134 experimental videos of an approaching vehicle from a pedestrian curbside perspective. On each trial, the vehicle either began decelerating (with one of three onset distances) or maintained speed. Across four sequential blocks, the prevalence of FBL-equipped vehicles increased from 0% to 83%. The session ended with a trial in which the FBL activated without deceleration. Participants pressed a key as soon as they perceived braking. A generalized linear mixed-effects model showed that deceleration was detected substantially faster when the FBL was present. Critically, an FBL × block interaction indicated that braking detection for vehicles without an FBL slowed in the (last) high- prevalence block, despite earlier practice-related improvements, suggesting increasing reliance on the explicit cue as it becomes common. In the oddball trial, all participants responded after FBL activation, indicating strong cue reliance. Eye-tracking results supported this interpretation. Fixation durations peaked around braking onset, indicating intensive processing; when the FBL was present, fixation durations around onset and immediately thereafter were reduced, consistent with faster extraction of braking information and lower attentional demand. Together, the findings suggest that increasing exposure to an eHMI can change how road users interpret vehicle state itself (e.g., whether it is braking), and that these effects depend on how prevalent the eHMI is. The absence of an eHMI may increase uncertainty or delay responses to vehicles without it, but such costs may still be outweighed when the eHMI is widely deployed.     «

Over recent years, there has been a lively debate about whether (automated) vehicles should be equipped with novel external human-machine interfaces (eHMIs) to support interactions with vulnerable road users. Explicit signals, such as a frontal brake light (FBL) that communicates vehicle deceleration to road users in front of the vehicle, have been shown to substantially speed up deceleration recognition. However, emerging evidence suggests that introducing such signals may also produce unint...     »