Design and Evaluation of a Smartphone-based Alarming System for Pedestrian Safety in Vehicular Networks

This paper proposes a depth-based alarming service for pedestrian's safety in streets. This service is based on attentional network in cognitive neuroscience field. For the detailed study, we developed an Android alarming App, letting a smartphone user be informed of risk in advance. The alarming App was evaluated with six types of alarm along with a remote control car. During experiment, four metrics were measured, such as response time, collision, disturbance, and satisfaction by recording and questionnaire. The results show that, among six sorts of alarm, providing pre-warning with colorful background was useful to a pedestrian's response time to a main warning, but transparency for background color was not useful. These results demonstrate that offering pre-warning alarm makes the users promptly avoid the collision with cars. On the part of preference, going against our assumption, people prefer an apparent warning to a transparent background-color warning for less disturbance. Participants expressed that the less clear notification message is provided with transparent background, the more disturbed they are. Through experiments, it is shown that the proposed two-level depth-based alarming service can significantly reduce a pedestrian's reaction time to a main warning message, leading to the provisioning of better safety for pedestrians.