Autonomous driving will become one of the mainstream modes of transportation in the future. Consequently, potential traffic safety issues associated with autonomous driving have received significant attention. Some researchers argue that introducing autonomous driving to vehicles may reduce the overall frequency and severity of crashes and personal injuries [1]. However, if misunderstanding occurs in the communication between autonomous vehicles and pedestrians, then undesirable consequences are inevitable [2]. Currently, drivers typically communicate their driving intentions to pedestrians via signals from the vehicle, eye contact, gestures, etc. [3,4,5,6,7]. These communication methods (facial expression, eye contact, gesture, vehicle movement, and the sound from the vehicle) allow pedestrians to clearly understand the intentions of car drivers and to be aware of upcoming vehicles [8,9,10,11,12]. While crossing a road, pedestrians can assess whether they can safely cross an intersection based on the speed and acceleration of vehicles as well as the distance between the vehicles and themselves [13,14]. However, unlike humans, autonomous vehicles have not yet developed the capability to communicate with pedestrians through using and interpreting implicit and explicit communication signals [15]. Furthermore, autonomous vehicles cannot accurately simulate the typical human–vehicle interactions, such as gestures and eye contact [2]. Therefore, traffic accidents and serious injuries may occur when the signals provided by autonomous vehicles to pedestrians and other vehicles are ambiguous. To ensure the safety and acceptance of autonomous vehicles and the trust of other pedestrians, communication strategies between autonomous vehicles and other pedestrians must be developed [16]. These strategies must ensure that autonomous vehicles can interact and communicate with other pedestrians for safer driving. This task is particularly challenging for mixed traffic involving autonomous and manual vehicles [17]. In particular, considering the most recent development of autonomous vehicles, informing about “vehicle behavior” and “vehicle intention” without any direct interaction between the driver and the external environment is becoming increasingly more important [18]. When autonomous vehicles begin to extend from the original restricted areas to open roads with both pedestrians and vehicles, the conventional vehicle indicator light and other typical interaction methods must be adapted appropriately. Researchers have investigated the effect of the intention transmission of autonomous driving from the visual level via lighting, light-emitting diodes (LEDs), or projections [2,14,15,19]. Additionally, researchers have focused on auditory feedback [14,19] and used additional devices [14] or humanoid robots [8] to assist in transmitting vehicle information. Pugliese, Brian J., et al. (2020) [20] reported that pedestrians rely primarily on visual signals to assess the safety risks when crossing a road, whereas they regard auditory signals as auxiliary information. However, a consensus regarding the optimal combination of light and sound for vehicle intention transmission has not been reached. Moreover, the factors of light and sound require further investigation. The optimal combination of light and sound that would result in the highest usability and is most consistent with the cognitive preferences of pedestrians is yet to be identified.

The purpose of this study is to investigate the signal design of autonomous driving based on different scenarios, such that important factors that affect cognition can be identified and a better design combination that is conducive to the cognition of pedestrians can be determined. The findings of this study will serve as a reference for future interactions between pedestrians and autonomous vehicles.