This proposal is to collect perishable data on the physical response of the transportation infrastructure in New York City following Hurricane Sandy. It makes use of a new human-in-the-loop smartphone-based crowd-sourcing sensing technology, called TrafficTurk. TrafficTurk is a smartphone application which enables intelligent, human?centric sensing of traffic flows during extreme events.
The aftermath of Hurricane Sandy represents a rare opportunity to observe transient behavior of a transportation network in response to a significant loss of physical infrastructure (due to flooding and gas shortages) and cyber infrastructure (due to loss of power for traffic control devices). The data gathered by this project, which will be shared with researchers across the country, will enable study of how traffic dynamics evolve after a major disruption to the cyber and physical components of a transportation infrastructure system. Potential benefits include improved preparedness and response to future disasters.
This project collected perishable data on the response of the roadway transportation infrastructure in New York City following Hurricane Sandy in October 2012. It used a new smartphone based traffic sensor called TrafficTurk to collect detailed traffic information at selected intersections in New York City. It also collected coarse GPS data from New York City taxis to understand citywide traffic conditions. Combined, the two data sources provided a way to measure the resilience of the transportation network after Hurricane Sandy caused a significant loss of critical physical infrastructure due to flooding and cyber infrastructure due to an electricity outage in lower Manhattan. The collected data was analyzed to understand the congestion before, during, and after Hurricane Sandy. A method was developed to quantify the magnitude of the disruption compared to typical traffic conditions. Key findings include i) the traffic conditions throughout most of the city returned to typical congestion levels about 6 days after the storm, and ii) The evacuation prior to the storm caused relatively minor disruption compared to the extreme disruption immediately following the storm. The data collection and analysis approach developed in this work is effective for measuring city-scale traffic dynamics after a natural disaster. While there may be no perfect strategy for moving people efficiently on damaged infrastructure, managing the traffic during the recovery may provide opportunities to reduce congestion levels following disruptive events.
