Poster: Visual Cue-Based VRU Protection on Smartphones

As autonomous vehicles loom as reality, and the vehicle communication starts to be enforced by law from 2020, protecting vulnerable road users (VRUs) using vehicle communication is receiving attention. The current vehicle-to-pedestrian (V2P) communication as stipulated by the standards such as SAE J2735 implies that it is the vehicles that take the responsibility for VRU protection. User devices are essentially beacons that transmit Personal Safety Messages (PSMs), and upon receiving PSMs, the drivers (or autonomous vehicles) take necessary measures to protect them. We, however, believe that the road users also need information about nearby vehicles to protect themselves from dangerous situations. Using other technologies than the standard Dedicated Short Range Communication (DSRC), there have been existing works for VRU protection. They use Wi-Fi or Wi-Fi Direct as replacements of DSRC. An automaker tested DSRC for VRU protection, but no technical detail has been presented. An important issue with the existing VRU protection proposals is that they are fraught with false alarms, which lowers the utility of the whole idea. Although one can come up with a highly precise collision prediction model, any such model will generate a huge number of false alarms, especially in urban environments. For example, if a pedestrian walks along a sidewalk well protected from the driveway, all passing vehicles will generate an alert to the pedestrian and vice versa. So our approach instead provides intuitive visual cues to the smartphone user looking at the screen, so that they can use their discretion to determine the level of danger for themselves.