2026-09-09 バージニア工科大学（Virginia Tech）
Zoox team member Daniel Edillor demonstrates a crosswalk test scenario designed to measure participant response times to audio tones played by the approaching test vehicle to signal it will yield. Photo by Erem Memisyazici for Virginia Tech.
＜関連情報＞
- https://news.vt.edu/articles/2026/09/vtti-zoox-automated-vehicle-pedestrian-audio-study.html
- https://static.nhtsa.gov/esv/pdf/ESV/Proceedings/28/ESV26-295.pdf
レベル4自動運転システムと歩行者の相互作用： 新しい聴覚型外部ヒューマン・マシン・インターフェースの評価 PEDESTRIAN INTERACTIONS WITH LEVEL 4 AUTOMATED DRIVING SYSTEMS: EVALUATION OF A NOVEL AUDITORY EXTERNAL HUMAN MACHINE INTERFACE
Gabrial T. Anderson, Srinath Sibi, Sadie Cooke, Jeremy Yang, Lowell Pickett, Ellen Xing, Kyle Evans, Jasmine Kim, Huadong Wang, Rossitza Kotelova, Rafael Patrick, Tanner Upthegrove, Stephen A. Ridella, Luke Neurauter, & Sheila G. Klauer
ABSTRACT
With the advent of SAE Level 4 Automated Driving Systems (ADS) there is no longer a human driver to provide supplemental cues to pedestrians. To replicate this driver-to-pedestrian communication, ADSs may need additional external methods of communication to aid pedestrians and other road users with understanding and anticipating vehicle actions. The current study examined the effect of novel sounds on pedestrian behavior. Forty participants were balanced across gender and vision status (i.e., 20 low-vision/visually impaired and 20 normal sighted participants). Participants experienced two scenarios: (1) a jaywalking scenario and (2) a pedestrian crosswalk scenario. In each scenario, an external speaker on the automated vehicle played non-speech-based, auditory sounds to communicate with participants. Participants provided objective crossing decisions and subjective ratings about the sounds they experienced. For the jaywalking scenario, time to not cross (TTNC; i.e., time to decide “I will not cross the road”; faster is better) showed that the effect of sound was significant as compared to the baseline vehicle sound. Further, all test sounds resulted in similar TTNC metrics as the horn sound. This suggests that the test sounds were as effective as the horn sound in influencing TTNC for jaywalking scenarios. For the pedestrian crosswalk, results suggested that sound did not influence crossing decisions. This suggests that for pedestrian-ADS interactions at crosswalks, repeated sound exposure may be needed for learning what a sound is communicating or that additional external communication may not be as necessary above that of the overall motion of the vehicle.
