University of California Institute of Transportation Studies

Automated Driving Systems (ADSs) are developing at a rapid pace and even testing on public roads, but pedestrians’ interaction with ADSs is not comprehensively understood and investigated to ensure safe operations of ADSs. The objective of this study is to investigate the effective interaction between ADSs and pedestrians. We developed prototype interfaces using different modalities, for example text vs. symbol and variety of symbols. These interfaces communicated three types of information: 1) intention of ADS; 2) instructing pedestrians what to do, and 3) ADS’s awareness of pedestrians. We tested the interfaces through two field studies in three uncontrolled intersection with crosswalks. The Wizard of Oz method was used, in which an experimenter worked as a driver and was invisible by wearing an outfit to simulate an automated vehicle (AV). The interfaces were displayed on an LED panel mounted on the AV. Results showed external interface on the AV didn’t change the decision time for pedestrians to cross. However, the vehicle movement patterns (e.g., slowing down the vehicle speed) continued to be a significant cue for pedestrian. All participants perceived the communication of the ADS’s intent (e.g., “stopping” printed on the LED panel) and the advisory information from the ADS (e.g., an icon that indicated it was safe for pedestrians to cross); these were both more effective than trying to convey the awareness of the ADS (e.g., an icon with an open or closed eye). The subjective ratings showed positive effects of the interfaces that were easy to understand (e.g., text interface and symbol interface) as they did help pedestrians to feel safe and trustful when interacting with the ADS.