A car whose driver is on their phone will often be slow, and wander in its lane. If they’re about to make a right turn, you can usually tell by the way the car is positioned even if the driver isn’t using their turn signals. Someone who’s getting ready to pull away from a stop will often inch forward a little before they actually get going.
#AVS VIDEO EDITOR TEXT BEFORE VIDEO DRIVERS#
I was thinking more for body language.Ĭars do have body language, and drivers telegraph their intent without realizing. I wasn’t thinking of it in terms of the car making eye contact with pedestrians in order to tell them it’s seen them, though. It’s not gonna happen, but it pretty evidently would make AVs safer for pedestrians. It’ll be fun to watch the car designers go nuts with this, anyway. Automated vehicles need to operate in the world of human culture and communication, and if that means cars will end up being designed with big, hilarious eyes, then, well, so be it. Part of making a truly automated vehicle will be to automate those innate communication processes we take for granted. We communicate almost constantly as we drive around other beings and people in other vehicles, whether we realize it or not. There’s a lot of interesting information in these papers and it’s focused on a part of AV development that tends to get overshadowed by the more dramatic actual driving and object avoidance and other parts of the automated driving problem. The team’s work is extensive, expanding to situations beyond just basic street-crossing situations to include more advanced, “critical” situations including ones with a hurried street crosser and the car both willing and unwilling to yield to the pedestrian: That last sentence there is especially interesting: gender differences? It does seem that the researchers did find differences in how men and women reacted to the eyes, which I wouldn’t have expected.įor example, they found eyes reduced accidents for men, while it improved efficiency for women: In addition, the results showed gender differences regarding critical and non-critical scenarios in AV-to-pedestrian interaction. The results showed that the eyes can reduce potential traffic accidents and the gaze directions can increase pedestrians’ subjective feelings of safety and danger. We conducted an empirical study using 360-degree video shooting of a real car with robotic eyes.
Thus, the pedestrian can judge that they should not cross the street, avoiding potential traffic accidents. If the car is not looking at the pedestrian, it implies that the car is not recognizing the pedestrian. In this work, we examine a more critical question “can eyes reduce traffic accidents?” In order to answer the question, we consider a critical street-crossing situation where a pedestrian is in a hurry to cross a street. A previous work showed that autonomous vehicles’ (AVs) eyes help pedestrians make faster street-crossing decisions. Various car manufacturers and researches have explored the idea of adding eyes to a car as an additional communication modality. There’s a lot of interesting stuff in there in addition to the real-world tests, they also conducted virtual reality tests, and in those found a 64% reduction of unsafe street crossings when the cars had eyes on them compared to those that did not. Here, they made a video demonstrating the project: The team acknowledged that equipment like turn indicators do some of this sort of work, visually conveying the car or driver’s intent, but it seems the eyes have potential to convey even more information. The team of University of Tokyo researchers called their project Gazing Car, which involved fitting a golf cart-sized car with large, robotic eyes that conveyed where the car’s attention was focused and the car’s intended general path. Ideally, big googly eyes, and, more importantly, they should be positioned in the correct place, where the headlights are located, and which is very much not the windshield. One team in Japan has a simple solution: give the cars eyes. Cars, even automated ones, lack human-style eyes, which makes this sort of communication impossible. Our eyes telegraph a hell of a lot about our focus, attention, and intentions. This communication can take the form of body language, motion, behavior, or, perhaps most powerfully, direct eye contact. The issue is that we, as humans (or possibly bonobos) rely far more than we realize on unspoken communication with other drivers, pedestrians, cyclists, even animals. One of the most under-appreciated problems with self-driving cars isn’t really a technological problem at all it’s a social and communication problem.
0 kommentar(er)
