New York City Skyline

Spotlight on COVID-19

  • A new research paper by NYU CUSP and NYU Wagner Clinical Assistant Professor of Urban Planning and Public Service Neil Kleiman outlines in light of COVID-19 how a growing network of foundations, universities and advocacy groups can advance new inter-governmental arrangements that can better support cities and local government. Professor Kleiman’s research was also featured in an article by Inside Philanthropy.
  • New research by Associate Professor Quanyan Zhu simulates panic-buying behavior with the aim to provide authorities with information needed to adequately allocate and provide resources.
  • A new paper by Professor Oded Nov and Assistant Professor Rumi Chunara provides data on the feasibility and impact of video-enabled telemedicine use among patients and providers and its impact on urgent and non-urgent health care delivery from one large health system at the epicenter of the COVID-19 outbreak.
  • The GovLab recently released a living repository for data collaboratives seeking to address the spread of COVID-19 and its secondary effects. The repository invites individuals to share projects that show a commitment to privacy protection, data responsibility, and overall user well-being, and has identified and sourced over 220 projects thus far.

Research

  • The Behavioral Urban Informatics, Logistics, and Transport Laboratory (BUILT Lab), led by Assistant Professor Joseph Chow, will be working on several new grants, including:
    • “Urban Microtransit Cross-sectional Study for Service Portfolio Design” – a collaboration with Rae Zimmerman at Rudin Center, funded by C2SMART, this project will examine data from Via across multiple cities to develop a cross-sectional model of microtransit operations that can help cities manage portfolios of project options.
    • NYSDOT T.A.#SR-20-02 — “Statewide Mobility Services Program Strategic Procurement Planning” – an assignment from the on-call agreement that C2SMART has with NYS DOT, under which we will help them better understand the Mobility-as-a-Service initiatives around the country to prepare for their next statewide mobility service program.
    • “EAGER/Collaborative Research: Enable elastic capacity for transportation infrastructure through a transmodal modular autonomous vehicle system” -this NSF project  (CMMI-2022967) is in collaboration with Xiaopeng Li at University of South Florida to study how modular autonomous vehicles (e.g. https://www.next-future-mobility.com/) can provide more capacity flexibility to transportation systems; he will study the traffic flow aspects while my team will look into fleet operations.
  • NYU Tandon Assistant Professors Elizabeth Hénaff, Andrea Silverman, and Tega Brain, CUSP Research Assistant Professor Charlie Mydlarz and CUSP Smart Cities Postdoctoral Associate Junaid Khan were awarded a C2SMART grant to develop a street-level flooding sensor. This project is an offshoot of the ongoing urban flooding microbiome project that Professors Hénaff, Silverman, and Brain were funded through the Marron Foundation.
  • Industry Assistant Professor Benedetta Piantella was invited to co-host a Live YouTube show every Tuesday at 3pm EST about STEM and Engineering with Arduino co-founder Massimo Banzi called “Bar Arduino Worldwide.”
  • CUSP Executive In Residence Dr. Jurij Paraszczak led an IBM Journal of Research and Development issue on Disaster Management, with guest editors from IBM. The full issue can be found here.
  • The AI4CE lab led by Assistant Professor Chen Feng published a paper “SPARE3D: A Dataset for SPAtial REasoning on Three-View Line Drawings” in CVPR’2020. The paper presented a series of new visual learning tasks, i.e., spatial reasoning on line drawings, which originates from the civil/mechanical engineering domains. The paper revealed some surprising results that state-of-the-art convolutional networks’ spatial reasoning performance in SPARE3D is almost equal to random guesses, although these networks achieved superhuman performance in many other visual learning tasks like object detection and segmentation. This paper aims to stimulate new visual learning research for machines to understand and reason about 3D geometric data to advance the intelligence level in traditional engineering applications. The research is supported by the NSF CPS program under CMMI-1932187 (https://nsf.gov/awardsearch/showAward?AWD_ID=1932187). The source code and data are released for education and research purposes online: https://ai4ce.github.io/SPARE3D.
  • A new study by NYU CUSP and NYU Steinhardt Assistant Professor Ravi Shroff and his colleagues at the Stanford Open Policing Project, found that in a dataset of nearly 100 million traffic stops across the United States, black drivers were about 20 percent more likely to be stopped than white drivers relative to their share of the residential population.
  • The GovLab recently launched a free MOOC on “Collective Crisis Intelligence.” In collaboration with 11 partner institutions around the world, the course is open to anyone, and designed to help institutions improve disaster response through the use of data and volunteer participation. You can watch all the modules here and read more in this press release.
  • The GovLab Director Beth Simone Noveck, who also serves as chief innovation officer for the State of New Jersey, is collaborating with Courant Professor Lakshmi Subramanian to deliver data and predictive analytics to Governor Phil Murphy and multiple state agencies. You can read more about the collaboration featured in NYU Tandon’s “On the frontline” coronavirus updates here.
  • On April 21, The GovLab launched a new initiative with Microsoft, the Open Data Policy Lab, a resource hub supporting decision-makers working on accelerating the reuse and sharing of open #data for the benefit of society. You can check out the website here and read more about it in this blog post.
  • On April 30, The GovLab and UNICEF, as part of the Responsible Data for Children initiative (RD4C), shared a set of lightweight and user-friendly tools to support organizations and practitioners seeking to operationalize the RD4C Principles. These principles—Purpose-Driven, People-Centric, Participatory, Protective of Children’s Rights, Proportional, Professionally Accountable, and Prevention of Harms Across the Data Lifecycle—are especially important in the current moment, as actors around the world are taking a data-driven approach to the fight against COVID-19. Read all about them here.

Publications

  • Visiting Scholar Chaogui Kang had a new paper that has been accepted for publication:
    • Kun Qin, Yuanquan Xu, Chaogui Kang, and Mei-Po Kwan. “A graph convolutional network model for evaluating potential congestion spots based on local urban built environments.” Transactions in GIS, 2020.
    • Automatically identifying potential congestion spots in cities has significant practical implications for efficient urban development and management. It requires the ability to examine the relationships between urban built environment features and traffic con-gestion situations. This article presents a novel and effective approach for achieving the task based on a machine-learning technique and publicly available street-view imageries and points-of-interest data. The proposed multiple-graph-based convolutional network architecture can: (1) extract essential urban built environment features from street-view imageries and neighboring POIs; (2) model the spatial dependencies between traffic congestions on road networks via graph convolution; (3) evaluate the risk level of road intersections to emerging potential congestion situations based on the local built environment features. We apply the model to the Wuhan in China and predict the potential congestion spots across the city. The results confirm that the model prediction is highly consistent (about 85.5%) when compared to the ground-truth data based on traffic indices derived from a big taxi GPS trajectory dataset. This research enhances the understanding of traffic congestion situations under various geographic, societal and economic contexts based on easily accessible road, street-view and POI datasets at large spatiotemporal scales.
  • Mathieu Andreux, Tomás Angles, Georgios Exarchakis, Roberto Leonarduzzi, Gaspar Rochette, Louis Thiry, John Zarka, Stéphane Mallat, Joakim Andén, Eugene Belilovsky, Joan Bruna, Vincent Lostanlen, Muawiz Chaudhary, Matthew J. Hirn, Edouard Oyallon, Sixin Zhang, Carmine Cella, and Michael Eickenberg. “Kymatio: Scattering Transforms in Python.” Journal of Machine Learning Research. 21(60):1−6, 2020.
    • The wavelet scattering transform is an invariant and stable signal representation suitable for many signal processing and machine learning applications. We present the Kymatio software package, an easy-to-use, high-performance Python implementation of the scattering transform in 1D, 2D, and 3D that is compatible with modern deep learning frameworks, including PyTorch and TensorFlow/Keras. The transforms are implemented on both CPUs and GPUs, the latter offering a significant speedup over the former. The package also has a small memory footprint. Source code, documentation, and examples are available under a BSD license at https://www.kymat.io.

Education

It’s not too late to apply to CUSP – we extended our final application deadline! Apply by May 15th to join us in Fall 2020. Ready to finish your application?