Dynamic Transportation Network Vulnerability Assessment Algorithms

Principal Investigator: 
Performance Period: 
July 2015 to June 2016
Project Status: 
In Progress
Commercialization Status: 
N/A
Project Keywords: 
transportation network vulnerability
dynamic demand profiling
transportation network graph extraction
risk modeling
decision analysis
Abstract: 
The goal of this research is to develop dynamic transportation network vulnerability assessment algorithms. Two primary inputs are needed to execute these algorithms: (1) the static graph representation of a transportation network and (2) an accurate profile of the volume of traffic wishing to utilize links within the network throughout the day. Thus, to enable application of our algorithms to any city or region of interest, we will first implement a static map extraction tool to translate data contained in open source map software into a standard format used by the transportation engineering research community [B14]. To obtain dynamic demand profiles, we will develop a platform-independent smartphone application to collect anonymous data on travelers utilizing various modes of transportation, including automobiles, public transportation, and pedestrian flows. From this data, we will construct empirical time series models to characterize the congestion typically experienced at a given time and location within the network. We will also utilize Google traffic data if CREATE's Federal Coordinating Committee (FCC) or Scientific Advisory Committee (SAC) can facilitate access. The static graph representation of the network and travel demand time series will serve as inputs to game theoretic transportation network vulnerability assessment algorithms to identify the critical links and nodes of a network as a function of time. This time varying vulnerability assessment will be significantly more informative than the current state of the art [FRL], which is limited to static methods that cannot consider potentially rapid changes in vulnerability over time. Defenses allocation based on static vulnerability assessment may therefore only approximate an optimal strategy. Moreover, an accurate snapshot of transportation network congestion at the time of an incident will be essential to coordinate evacuation and emergency response.