Page 11 - Annual Report - interactive demo
P. 11
Quantifying Uncertainty and Distributed Adaptive Control for Unanticipated Traffic Patterns
Principal Investigator: Saif Jabari, NYU Abu Dhabi
Disruptions to city traffic flows are increasing, and these disruptions can cause hours of delay in busy urban networks. This research investigated the levels of pene- tration of CAVs needed to determine traffic states with sufficient accuracy for effective traffic management, when combined with data from stationary sensors.
The researchers also developed a new kind of network control algorithm, called position-weighted backpres- sure (PWBP), that can use data from CAVs to capture realtime traffic dynamics, allowing for adaptive signal control to efficiently clear congestion following an inci- dent. The researchers tested various incident scenarios in simulation, evaluating how well PWBP performed compared to other traffic control strategies. In all tested circumstances, PWBP was able to accommodate higher numbers of vehicles, reduce total network delay and congestion propagation speed, and improve the net- work’s recovery from heavy congestion and response to an incident.
Integrated Analytics and Visualization for Multi- Modality Transportation Data
Principal Investigator: Claudio Silva, NYU
This research project developed a data-driven approach for modeling cities, which plays a fundamental role in urban planning. Using a new data set composed of tens of millions of images from New York City, the research team presented a framework that leverages recent advances in computer vision to efficiently handle such a large collection of complex images and constructed
a map of relative pedestrian density in Manhattan. The researchers also developed an interactive visual analysis tool for the exploration of this large collection of images. Their approach computes a set of feature vectors for each image and makes use of a memory-efficient index to interactively answer queries about this data. Through a visual interface that couples the image data with other urban data sets, users can interactively query, explore and analyze visuals in a city over both space and time to virtually audit the built environment and help answer nu- merous questions about the complex system of cities.
Simulated test network in Abu Dhabi used to evaluate the Example usage of the Urban Portfolio visual analysis tool to performance of the PWBP algorithm. locate intersections without tactiles pavements.
C2SMART Center Annual Report 10