We make cities better places to work and play
The research center at the forefront of all things urban
C2SMART (Connected Cities with Smart Transportation), a USDOT Tier 1 University Transportation Center launched in 2016, uses cities as living laboratories to study challenging transportation problems and find data- and tech-driven solutions. Their research encompasses autonomous mobility, ride sharing and micro-mobility, infrastructure resiliency, transportation/traffic safety, big data for planning for smart cities, equity in transportation, accessibility for disadvantaged travelers, and more.
With a focus on accelerating technology transfer from the research phase to the real world, researchers at C2SMART are making state-of-the-art predictive artificial intelligence (AI) and machine-learning (ML) models, as well as high-fidelity transportation simulation models, to decision-makers to address a wide range of urban mobility problems and break down the barriers to innovation. Over the course of the last year, C2SMART also focused on the effects of the pandemic on urban mobility, given the unprecedented shift in demand and usage with changed travel demand, economic activity, and social-distancing and stay-athome policies.
The Center has deep connections to the city and several key initiatives. One of the most innovative and timely initiatives of C2SMART researchers was in developing a deep-learning object-detection method to use live video feeds from public Department of Transportation (DOT) cameras installed on city streets to determine pedestrian and traffic volume densities on city streets remotely using existing infrastructure. This effort, originally envisioned to estimate the ability for pedestrians to social distance, allowed researchers to also observe the drop and recovery of pedestrian, bicycle, and vehicular traffic volumes using existing infrastructure. This provided a new data source while also not exposing researchers to potential risk of transmission of COVID while collecting data. As part of this work led by Senior Research Associate Jannie Gao and Professor Kaan Ozbay, they used innovative AI-based approaches for image processing that have been widely recognized in the field.
C2SMART researchers Professors Hani Nassif and Kaan Ozbay, are working hand-in-hand with NYCDOT to address one of the most pressing infrastructure problems of the city by collecting real-time data using high-accuracy sensors on the cantilevered section of the Brooklyn-Queens Expressway (BQE). This section of the BQE, which was studied and reviewed by Mayor de Blasio’s Expert Panel, is a critical transportation link in New York City and the center of a multibillion-dollar rehabilitation effort. Through C2SMART’s work on the panel, a new approach to measuring and understanding the load on the structure was observed. Subsequently, C2SMART was formally added to NYCDOT’s team with real-time monitoring of the structure using new technologies, including high-tech weigh-in-motion (WIM) sensors. This pioneering research project has the potential to change the way weight limits are enforced on our aging and critical infrastructure.
These two efforts over the past year extend Professor Ozbay’s visiom when establishing C2SMART as a USDOT Center, and have helped C2SMART grow into a premier research center for solving important urban mobility and infrastructure problems using AI and ML-based data analytic and big data. C2SMART continues to also be a trusted research provider for the New York State Department of Transportation (NYSDOT), with research projects ranging from traffic management software to bridge deterioration, under their $12 million on-call partnership.
C2SMART’s FloodSense project was launched in 2020, led by (Elizabeth Henaff, Tega Brain, Andrea Silverman, and Charlie Mydlarz), and is part of the NYC FloodNet consortium, in partnership with CUNY and the NYC Mayor’s Offices of Climate Resiliency and the Chief Technology Officer (www.floodnet.nyc). The project developed a publicly accessible platform that provides real-time flood information through revolutionary sensors capable of overcoming common sensor challenges (including durability, connectivity, and power). The sensors are designed to help cities respond to emergencies by providing real-time information on flood depth, frequency, and duration. A data dashboard is currently under development to also allow community members to access flood data and additional data streams, such as rainfall data, 311 flooding complaints, and social media feeds, as well as the digital infrastructure necessary to log, process, and present this data for quick and efficient response.
“The New Normal: Combating Storm-Related Extreme Weather in New York City,” a recent report by the NYC Mayor’s Office, outlines a series of planned initiatives to improve the resiliency of New York City in light of recent Hurricanes Ida and Henri. Among them is a proposed expansion of the Flood Sensor Network developed by the FloodNet team citywide.
The emerging transportation modes and technologies we’re exploring have made an impact. The Center’s researchers created the first open-source multi-agent simulation model for New York City, called MATSim-NYC, to support agencies in evaluating policies such as congestion pricing — since measuring demand for emerging transportation technologies and policies can vary depending on a multitude of factors, and those effects are not well-captured with industry-standard static travel demand models.
Research by Joseph Chow, Deputy Director of the Center, recently published in the journal Transportation Science, involves development of an algorithmic model that optimizes Electric Vehicle (EV) charging for car-sharing fleets, solving a major challenge as society moves to EVs that are cleaner than internal combustion engine vehicles but take longer to fuel.
We’re also concerned with safety and accessibility. Professor Ozbay and his students are continuing to advance the field of traffic safety by taking advantage of the unique data obtained from connected vehicles and drones. As part of this work, in 2021 they published their traffic safety-related research in Analytical Methods in Accident Research and Accident Analysis and Prevention journals with a focus on better understanding the causal relationships among actual crashes and near-crash events with the goal of developing proactive safety models.
Professor Ozbay’s team is leading the task of field evaluation of a smart app for pedestrians with visual disabilities as part of the one-of-a-kind, multi-million-dollar USDOT NYC Connected Vehicle Pilot test led by NYCDOT.
Joint work between NYU and UTEP at the Center led by C2SMART Associate Directors Kelvin Cheu and Joseph Chow used machine learning to explore the possibilities of an algorithm called a recommender system — like the technology Amazon uses to suggest products, or that Netflix uses to recommend movies — for mobility-on-demand services. Public paratransit services, private rideshare companies, and future autonomous vehicle fleets could use them to improve operations and lower costs.
Our students are a priority. In the fall of 2020, C2SMART launched the “C2SMART Student Learning Hub,” free for all consortium member students. Students were able to access learning from a variety of course domains, including data science, computer science, and traffic simulation. The Hub is designed to offer students hands-on experience to learn the tools and skills they will need as they advance their careers, whether in academia, industry, or within government agencies.
Throughout Fall 2020 and Spring 2021, the Student Learning Hub offered a variety of courses, taught by a range of experts in the transportation field. This semester, we have attracted 105 students, across 14 universities, from 6 states in the US and 7 countries internationally. We worked with agency and industry partners to deliver programs and provided our students with access to researchers and professionals to learn both professional and academic skills. This initiative has been recognized by USDOT as a UTCleading effort to be adopted more widely.
C2SMART presents groundbreaking research at the TRB conference
To see the future of transportation, look no further than the 100th annual meeting of the Transportation Research Board (TRB). The event typically brings together 14,000 experts and practitioners from around the world to discuss emerging trends in transportation and infrastructure, with an eye to the future of how we get from point A to point B. At this year’s TRB, C2SMART once again made its mark. More than 100 C2SMART faculty, postdocs, and students gave presentations at the event, covering everything from how ridesharing technology affects mass transportation usage to the cultural perceptions of autonomous vehicles and how that affects their adoption.
Center for Urban Science + Progress
We make urban environments more livable using data and New York City as a living laboratory
If you’re looking to be part of datainformed, tech-driven solutions to urban problems, there’s simply no better place to go than Brooklyn — home to the Center for Urban Science and Progress (CUSP). Being fully embedded in the fastest-growing technology ecosystem in the country means there’s an unprecedented amount of data to work with and a plethora of ways to apply it in service of cities worldwide. Researchers at CUSP have a front-row seat to explore ways to make cities more resilient, more accessible, and more sustainable, while providing tools to city agencies and governments to evaluate how to make those changes happen in a smarter, more effective, and more equitable way.
- Let’s start with the fact that in a city of 8 million, there are often 8 million points of view, but one thing that seems to resonate for a vast majority of city dwellers is noise; and while the city that never sleeps in a badge we wear with pride, sometimes a good night’s sleep is a necessity. Noise pollution is the single most reported complaint to New York City’s 311 system and it is a concern across various city agencies, from the Department of Health and Mental Hygiene to the Department of Buildings. With the support of the NSF and City health and environmental agencies, CUSP — led by Professor Juan Pablo Bello, who also directs the Music and Audio Research Lab (MARL) at NYU Steinhardt — launched Sounds of New York City (SONYC), a first-of-its-kind comprehensive research initiative to understand and address noise pollution in New York and beyond. The project — which involves large-scale noise monitoring — leverages the latest in machine learning technology, big data analysis, and citizen science reporting to more effectively monitor, analyze, and mitigate urban noise pollution. The ultimate goals of the team, which includes Co-PIs Associate Professor Luke DuBois and Professor Oded Nov, are to create technological solutions that enable city agencies to take effective, information-driven action for noise mitigation.
Bello is an author on the paper “Deep Convolutional Neural Networks and Data Augmentation for Environmental Sound Classification,” which recently won the Institute of Electrical and Electronics Engineers (IEEE) Signal Processing Society (SPS) Signal Processing Letters Best Paper Award. In it, they set forth the idea of employing deep convolutional neural networks (CNNs) — a class of neural networks originally developed for computer vision tasks — to learn discriminative spectro-temporal patterns in environmental sounds. They were among the first researchers ever to explore this application.
- In other CUSP news, Associate Professor of Urban Analytics Daniel B. Neill — who is also an associate professor of computer science and public service at NYU’s Robert F. Wagner Graduate School of Public Service and the Courant Institute Department of Computer Science — is leading a three-year research project centered on the growing use of AI by urban, public sector organizations, work that will include the creation of open-source tools for assessing and correcting biases. The end result, his team hopes, will be to reduce incarceration by equitably providing supportive interventions to justice-involved populations, to prioritize housing inspections and repairs, to assess and improve the fairness of civil and criminal court proceedings, and to analyze the disparate health impacts of adverse environmental exposures, including poor-quality housing and aggressive, unfair policing practices, thus advancing social justice for the most vulnerable among us.
Debra Laefer honored for work tracking COVID-19 spread
This year Professor of Civil and Urban Engineering and CUSP faculty member Debra Laefer was honored by the New York State GIS Association, a group of Geographic Information Systems professionals, for her use of geospatial data in 3D-mapping the spread of Covid-19. Her study also set the groundwork for new machine-learning models to speed the analysis of how a virus spreads in urban areas.
We make cities safer when disaster strikes
NYU Tandon researchers are at the forefront of urban resiliency. Consider what happens when a transformer overloads — a not uncommon occurrence.
Because an overload in one area may cripple services in distant neighborhoods, full knowledge of subsurface infrastructure is critical to improving natural hazard resilience, identifying threats and vulnerabilities, and designing effective mitigation strategies. CUSP faculty member and Professor of Civil and Urban Engineering Debra Laefer is now bolstering the ability of first New York City and, ultimately, of cities nationwide to prepare for and respond to crises and disasters by making critical information on community infrastructure robust, open, transparent, and easy for key stakeholders to share and act upon. Her latest project, called Unification for Underground Resilience Measures (UNUM), aims to create seamless, interoperable, 3D data storage and visualization systems — and encourage stakeholders to share data with appropriate security measures in place.
We make the power grid stronger, safer, and more efficient
Assistant Professor of Electrical and Computer Engineering and Goddard Junior Faculty Fellow Yury Dvorkin is conducting research that lies at the heart of the national debate over climate change and energy security — areas with enormous and direct societal implications. He explains that as a researcher, he feels responsible for taking every meaningful step possible to contribute science-informed input to that debate and inform both decision-makers and the general public about evolving perils — and the best strategies to address them.
His work revolves around the mathematical modeling of energy systems and the development and application of control, optimization, and system theoretic methods for their operations and planning. It’s a complex undertaking: accurately representing physics and economics of the power grid requires performing computations over largescale, hierarchical, multi-agent networks, and furthermore, as power grids adopt more customer-end and renewable generation technologies, it becomes more important to account for their inherent uncertainty and lower controllability, which in turn complicates modeling and computational efforts even more and impedes capturing the economic value of emerging smart grid technologies. It is also important to consider how these smart grid technologies can help achieve ambitious goals in other sectors, for example transportation electrification. As more electric vehicles are adopted in the U.S., it grows increasingly vital to ensure that every electric vehicle driver has access to reliable and affordable charging infrastructure. Dvorkin, also a faculty member at Tandon’s Center for Urban Science and Progress, is dedicated to ensuring that the complexity of integrating new smart technologies and legacy infrastructure doesn’t lead to more operationally conservative, less efficient, less reliable, and more wasteful practices. Among his work has been his creation of a stochastic and risk-informed market design, which cost-effectively accommodates the stochasticity of renewable generation resources and provides adequate incentives to market participants — a boon since traditional deterministic market designs face deteriorating efficiency as more renewable generation resources are deployed. He is also well-known for his research on smart electric power distribution, which has helped grid operators control and manage electric power distribution processes by carefully evaluating power availability and deliverability in the presence of “smart” demand response and distributed energy resources, while concurrently using better methods to achieve optimal power flow and operate grid infrastructure under uncertain conditions.
Dvorkin is often called to comment on power issues before government committees and in the media, and he has deep expertise in answering questions of where and in what quantities to roll-out emerging energy technologies to ensure their most efficient usage, maximize their complementarity, and reduce overall capital investment. Answering these questions, he explains, requires bridging the divide between policymaking and the economic-engineering modeling of power system expansion, especially as the nation embarks on an ambitious deep, society scale decarbonization under the leadership of the current administration. In the U.S., large-scale infrastructure systems like power grids often cross state jurisdictional boundaries, and Dvorkin is credited as the first researcher to present a model to account for these split jurisdictions, calling upon game theory and high-performance algorithms.
Among his most recent work is investigating emerging vulnerabilities, both engineering and social, that arise in electric power systems as they increasingly rely on new information and communication technologies — and are increasingly exposed to high-impact disturbances such as cyberattacks, natural disasters, and epidemics. He has, for example, developed a modeling framework to analyze cybersecurity threats that arise in electric power distribution systems with a sufficient number of “smart” resources, such as electric vehicles, and he is the first to carry out simulations of such attacks and evaluate their impacts using real-world data. Alarmingly, Dvorkin’s group has discovered that exploiting a vulnerability in smartphone apps for charging stations, malicious actors need to compromise only about 500 vehicles to cause a brown or blackout in Manhattan, a finding he immediately conveyed to ConEdison. He also recently ascertained that the COVID-19 pandemic was exacerbating socioeconomic disparities in access to urban infrastructure and quality of electricity service and designed a real-time dashboard so emergency responders could prioritize the most vulnerable communities.
Zhong-Ping Jiang elected to Academia Europaea
Professor of Electrical and Computer Engineering Zhong-Ping Jiang, who directs Tandon’s Control and Network (CAN) Lab, has been elected to Academia Europaea (The Academy of Europe). Jiang — who is also an affiliated member of Tandon’s Center for Advanced Technology in Telecommunications (CATT) and C2SMART — focuses on interdisciplinary problems that cross the borders between wireless communications and urban issues to advance connected and autonomous vehicles, robotic networks, cyber-physical systems, and more. He was elected as a foreign member of Academia Europaea on the basis of his outstanding contributions to the nonlinear small-gain theory and a theory for learning-based control being developed at Tandon.