Urban Science IntensivePrevious USI Projects

Class of 2016

 

WiFind Project: WiFi Mapping

CUSP Students: Alan Polson, Dan Quasney, Jeremy Neiman, Manushi Majumdar, Tianyi Gu, and Yunong Cao

CUSP Mentors: Charlie Mydlarz and Justin Salamon

USI Sponsor: CUSP Sounds of New York City (SONYC)

The WiFind Project has created a framework for collecting and analyzing the presence and strength of Wi-Fi networks. This data is crowdsourced via a custom-built mobile application and visualized on our website as a map.

WiFind can help inform a host of stakeholders including non-profit organizations striving to bridge the digital divide, city agencies charged with providing free Wi-Fi in public spaces, Internet of Things (IoT) enthusiasts searching for the best locations to install their internet-connected devices and researchers looking to examine the extent of Wi-Fi coverage in a city.

We have several data collection initiatives planned or in progress, which have already gathered over 15 million rows of data. We also provide the capability to access the raw data while ensuring privacy protection of any personal data that may be linked back to individuals.

The App and Website are live and thousands of rows of data are being added daily. The team invites you to visit the WiFind Project website at www.wifindproject.com.




Socio-Economic Characteristics Analysis

CUSP Students: Juan Mora, Diego Garzon, Nikhil Kishore, and Yanchao Xu

CUSP Mentor: Stan Sobolevsky

USI Sponsor: NYC 311

The project team will collaborate with NYC 311 in order to determine the relationship between socio­economic characteristics of New York City population (both residents and workers), and the number and types of service requests received by 311. The project will include the development of a web­-based or mobile visualization tool that can aid the agency to assess the relationships and metrics of their service and the type of complaints received within a defined geographical area, a machine learning based predictive and classification model, and a technical report summarizing the findings about the relationship between each type of service request and the socio­economic profiles of each area in New York City.




Consumer Fraud Detection

CUSP Students: Maxwell Feinglass, Svarmit Pasricha, Tianqi Jiang, Xiaoge Wu, and Yifan Xie

CUSP Mentor: Ravi Shroff

USI Sponsor: New York State Office of the Attorney General (OAG)

This project, sponsored by the New York State Office of the Attorney General (OAG), develops a tool that uses government and social media data to generate leads for consumer fraud investigations and to inform investigators about fraud patterns in various sectors of the economy. The tool consists of an interactive visualization dashboard linked to a database, and a natural language processing algorithm to filter social media data.




Illegal Dumping

CUSP Students: Calvin Li, Carlyle Davis, Lily Fung, and Tania Mazariegos

CUSP Mentor: Brendan Reilly

USI Sponsor: New York State Office of the Attorney General (OAG)

The team will work with the OAG to develop a predictive model to identify instances of illegal dumping of construction and demolition debris. The project aims to detect both waste creators and haulers who do not comply with city regulations.

By connecting various datasets on building sites, construction projects, complaints, and licensed waste haulers throughout the City, the team will establish extensive landowner and waste hauler profiles to identify the probability of illegal dumping.




Commercial Development

CUSP Students: Ding Ma, Erwan Lecun, Sarah Welt, Zeyu Jiang, and My Phan

CUSP Mentor: Stan Sobolevsky

USI Sponsor: NYC Economic Development Corporations (NYCEDC)

The project will develop a data ­driven approach to determining opportunities for viable commercial office space development in the outer­boroughs. This includes defining approximately ten indicators that can be used to highlight potential for commercial growth, including leveraging information like MTA turnstile counts, NYC DOB commercial building permits, and trends in employee density. The project will also translate the indicators into actionable insight through visualization and mapping.




Planning for Strategic Neighborhood Growth

CUSP Students: Avigail Vantu, Boyeong Hong, Diogo Miura, Lucy Kang, Vipassana Vijayarangan, and Yuan Lai

CUSP Mentor: Logan Werschky

USI Sponsor: NYC Department of City Planning

This project seeks to create a common operational picture for all neighborhood planners, enabling access to authoritative datasets aggregated from several City agencies and providing the ability to generate common metrics. The new metrics will facilitate the creation of standard neighborhood reports and other documents generated from the same data with a set of metrics derived using standard methodologies and delivered is a uniformly formatted document. The associated tool will save planners time and encourage the incorporation of more data analytics into the capital planning process. Furthermore, by having all planning divisions utilize the same data and metrics, planners within DCP and across agencies can then better coordinate and collaborate effectively.




Parks Condition Assessment

CUSP Students: Dhia Barnes , Michael Evans, Neil D’Souza, Venkat Motupalli, and Xia Wang

CUSP Mentor: Greg Dobler

USI Sponsor: NYC Department of Parks and Recreation

The team will conduct analysis and develop a model that describes the relationship between operational inputs and park conditions, provide proposed cleaning schedules for properties in Parks’ portfolio, and create data visualizations of key operational information.




Crime Patterns

CUSP Students: Linda Li, Lucas Chizzali, Philipp Kats, Siying Zhang, and Yi Zhang

CUSP Mentor: Martin Jankowiak

USI Sponsor: New York Police Department

The New York Police Department is implementing a new pattern detection algorithm. Patterns are crimes committed by the same individual or group of individuals using a similar Modus Operandi (MO). Patterns are a valuable target for enforcement and investigatory efforts. Identifying patterns from the sea of data and reports that are generated is challenging. This project will provide automated suggestions for similar crimes, thus helping detectives discover patterns more efficiently.




Taxi Demand Prediction

CUSP Students: Bhagwat Singh Bisht, Kiran Venkata Palla, Linfeng Zhou, Sachin Verma, Yining Fan, and Yuxiang Zhang

CUSP Mentors: Huy Vo and Kaan Ozbay

USI Sponsor: NYC Taxi and Limousine Commission (TLC)

This project will predict demand for legal street hail service across all neighborhoods of the City and will allow the TLC to create sound policy, and guide existing programs in the right direction. The team will create a model predicts passenger demand by NYC block or census tract, and visualizations that show how well demand has been met and is currently met.




Measuring Bus Reliability In NYC

CUSP Students: Bonan Yuan, Jiaxu Zhou, Matthew Urbanek, Sara Arango Franco, and Yuqiao Cen

CUSP Mentors: Huy Vo and Kaan Ozbay

USI Sponsor: NYC Department of Transportation (NYC DOT)

This project aims to develop a methodology for estimating bus travel times and their distribution as a function of relevant factors. In the first phase of the project, the team will develop algorithms to measure or estimate certain types of events associated with bus vehicles’ trips using publicly available data, such as departures from terminals, vehicle arrivals at stops, interruptions between stops, and customer demand at each stop. In the second phase, the team will analyze the resulting data and calculate metrics related to the performance of the buses with respect to their planned schedule and customers’ perceptions of dependability. Finally, in consultation with the DOT, the project team will develop hypotheses and test the significance of contextual factors on those performance metrics.




Quantified Communities

CUSP Students: Alejandro Porcel, Clayton Hunter, Eren Con, Jonathan Grundy, Maria Ortiz, and Tengfei Zheng

CUSP Mentor: Constantine Kontokosta

USI Sponsor: CUSP Quantified Communities

As part of the CUSP Quantified Community research initiative, this project explores the application of urban sensors to understanding neighborhood dynamics and how to effectively communicate information. The first phase of the project involves designing, calibrating, and deploying an array of environmental sensors around Red Hook, Brooklyn. The second phase involves analyzing the results of the sensor data, combining it with baseline demographic and city data, and communicating it to the residents of Red Hook through in-person citizen science days and Google Physical Web devices. Results of this project show that sensors can be useful in better understanding how neighborhoods operate and how to detect issues, and that residents of a neighborhood are interested in the data and how it can help them better their community.




Computer Vision

CUSP Students: Alec Mclean, Bilguun Turboli, Boya Yu, Denis Khryashchev, Priya Khokher, and Richard Nam

CUSP Mentor: Greg Dobler

USI Sponsor: CUSP Urban Observatory

This purpose of this project was to show that the field of computer vision can be used to count pedestrians. Using three different computer vision techniques, Blob detection, Histogram of Oriented Gradients, and Convolutional Neural Networks, the team will build models that will count pedestrians in real time. There are numerous benefits to incorporating computer vision into the way Cities count pedestrians. The benefits can be broken down into 3 categories: real- time information, autonomous extraction, and accurate counts.




Data Poverty

CUSP Students: Ramda Yanurzha, Jiheng Huang, Maria Filipelli, Adhlere Coffy, and Arno Amabile

CUSP Mentors: Dani Hochfellner and Logan Werschky

USI Sponsor: NYC Mayor’s Office of Data Analytics (MODA)

Data poverty is the situation in which one is deprived from the benefits of Open Data driven by the lack of access to, use of, and representation within data.

As an initial step to reducing data poverty, students will work with MODA and local community organizations to quantify data poverty in NYC. Specifically, they will identify factors contributing to data poverty, utilize a systems engineering approach and quantify the relationship between data poverty and its factors. In doing this they will refine the concept of data poverty, create a new methodology and index to quantify it, and developing a visualization to demonstrate differences across NYC communities to drive targeted outreach and action by the City.