The human element will constitute a critical component of tomorrow's Cyber Transportation Systems (CTS), either as drivers interacting with wireless messages in connected vehicle settings or as driver-passengers when vehicles are automated. Given that no engineered system will work perfectly all of the time, this raises important questions relating to the nature of the interaction between CTS and the human driver. Whereas the focus of the current literature has primarily been on hardware and software, in this pioneering project the PI and his research team will instead explore human-CTS interaction as the central design consideration. In accordance with this vision, they will develop and experimentally validate an innovative computational framework (mathematical model) for CTS that quantifies the effects of system design parameters on the key elements of human cognition and performance, which in turn will enable the derivation of design requirements and optimal parameter settings for connected and automated vehicles in order to maximize human safety.
For connected vehicles the model could be employed to determine minimal notification requirements (e.g., message transmission reliability) so as to maintain driver alertness and minimize driver distractions. For automated vehicles, the framework could be utilized to derive the minimal acceptable system reliability, optimal re-practice to minimize motor skill decay, and optimal lead time and design of the messages informing a driver to take over control of the vehicle (machine to human driver hand-over). Project outcomes will provide researchers in the field with the first tool to understand the fundamental mechanisms of human-CTS interaction (e.g., the impact of system reliability on human acceptance and trust), by predicting the effects of technology on driver behavior. The software will be open source, so that CTS researchers can easily modify their designs and/or explore the effect of modifying existing features and/or adding new ones as changes in technology may warrant.
