User Recognition at Mid-Block Crossings via Connected Vehicle Technology: An Evaluation of Driver Awareness via Eye Tracking and Stated Preference Data

Author:
Angulo, Austin, Civil Engineering - School of Engineering and Applied Science, University of Virginia
Advisor:
Smith, Brian, Civil & Env Engr, University of Virginia
Abstract:

Mid-block crosswalks prove to be confusing and dangerous for both drivers and pedestrians since users must rely on visual communication for safe use and operations. Using connected vehicle technologies, pedestrians can directly communicate with drivers, beyond using body language, by broadcasting safety messages from their mobile phone that are received by drivers in-vehicle. This study consisted of designing, developing, and deploying a cyber-physical pedestrian to vehicle communication system at a mid-block crosswalk to analyze drivers’ reactions to in-vehicle advanced warning messages, the impacts of in-vehicle advanced warning messages on driver awareness, and drivers’ acceptance of this technology. The application was designed to create a virtual advanced warning system at the mid-block crosswalk by geofencing designated areas in which users will be able to interact with each other via their smartphone or tablet.
In testing human subjects with, and without, advanced warning messages upon approaching the mid-block crosswalk, subject reaction, acceptance, and eye tracking data was collected. After analyzing the data from these performance metrics, it was found that the advanced warning message increased the rate at which drivers stopped for the pedestrian by a statistically significant 20%, increasing the yielding rate from less than half (48%) to well over half (68%). Furthermore, it was found through stated driver observations that 82.5% of participants found the application increased their awareness, 90% of drivers did not find the application more distracting than helpful, and 83.75% of drivers thought that the application was a technology they would like to see deployed in similar GPS navigational systems. Lastly, it was found that, in general, drivers statistically did not look at the pedestrian any more or less during their full exposure to the pedestrian; however, drivers looked at the pedestrian approximately 15-20% less after having just received the message at between the range of 260 to 200 feet away from the crosswalk. The results shown in this 260ft-200ft range were found to be significant at the 95% confidence level. Collectively, these results suggest that drivers were more aware of the pedestrian and their intent to cross the mid-block crosswalk with the advanced warning message.

Degree:
MS (Master of Science)
Keywords:
Safety, Connected Vehicle Technology, Cyber-Physical systems, Pedestrian, Mid-block
Language:
English
Rights:
All rights reserved (no additional license for public reuse)
Issued Date:
2018/03/12