The mission of CUSP is to develop data science tools to better understand cities and enable city agencies to best perform their duties. Some of those tools include sensor networks, citizen science data collections, and simulations.
Urban Observatory Control Room
The Urban Observatory deploys rooftop camera systems to continually collect visual, infrared, and hyperspectral imagery that reveals the urban “pulse” and detects anomalies. The UO also makes use of images flowing from a network of 600 street-level DoT traffic cameras. Applications span from assessing traffic safety interventions to identifying the use of energy efficiency technologies in or pollution emissions from individually identifiable buildings. The control room’s 28 screens bring the city to life from a data perspective. When Prof. Greg Dobler presented his work to TTI/Vanguard in February 2016, the Urban Observatory was in the proof-of-concept phase; it is now deployed across several sites in NYC.
Applications of motion capture technology go far beyond the realm of games, entertainment, and communication. Prof. Paul Torrens is interested in how people acquire local geographic information in uncertain and dynamic contexts. He uses motion capture to build agent-based models emphasizing nonverbal communication among people and environments in crowd scenarios. Autonomous vehicles will need to interpret just these types of crowd behavior. They will need to recognize pedestrian gestures (e.g., hailing cabs), differentiate attentive pedestrians from distracted ones, and respond appropriately to traffic cops’ hand signals.
Sounds of New York (SoNYC) Lab
Cities are noisy, but noise is transient. To enable more effective enforcement of New York’s noise ordinance, SoNYC is training algorithms to distinguish among the offending urban sounds New Yorkers call the city to complain about (construction noises, music, honking, dogs barking, and so on). The goal is to deploy a network of sensors and citizen science apps that can improve the odds that noise inspectors will be on location to measure major infractions when they occur.