Google Summer of Code 2025 at the City of Boston
Below, you will find a list of ideas we have for a Google Summer of Code contributors for the summer of 2025. Thank you for your consideration!
You can find guidance for applications and your project proposals here.
As our city navigates the rise of ride-sharing, delivery services, bike lanes, and evolving traffic patterns, the City of Boston’s Streets Cabinet faces increasingly complex challenges in optimizing our finite curb space. Traditional space-by-space planning approaches no longer suffice for managing what has become an intricate ecosystem of competing demands in our dense urban neighborhoods.
Boston’s transportation planners currently face a significant challenge. While planning street redesigns across the city’s diverse neighborhoods, they lack tools to understand and model how changes will impact the broader ecosystem. We need to prototype an interactive tool will help our planners to visualize and optimize curb usage in our neighborhoods. This tool will help city planners:
Future versions of the tool will:
We give this project a medium level difficulty. The project can be completed in 175 hours.
This project requires experience in web-based interactive mapping technologies and data visualization. The contributor should be comfortable working with spatial data and prototyping user interfaces with libraries like Bootstrap or Tailwind.
The mentors for the project will include key staff from the City of Boston’s Office of Emerging Technology, including one who served as a Google Summer of Code mentor at Code for America in 2011 and the City of Boston between 2021 and 2024.
Boston 311 serves as the City’s constituent service center. 311 is a non-emergency municipal service system that allows residents to report issues (like potholes, graffiti, noise complaints) and request city services through phone, web, or mobile apps. It serves as a centralized way for cities to track, manage and respond to citizen requests while keeping emergency 911 lines free for actual emergencies.
Over the years, we’ve developed various 311 dashboards to help residents and City officials understand service request patterns and response times. While these dashboards serve their purpose, we see an opportunity to create more engaging and interactive visualizations using modern web technologies that can efficiently handle large volumes of 311 request data.
This project will include the following activities:
Creating reusable visualization components using Mapbox GL and D3.js specifically designed for 311 service request data, including:
Implementing a component library with interactive features tailored for 311 data analysis:
Prototyping a 311 dashboard that demonstrates:
We give this project a medium level of difficulty. The project can be completed in 175 hours.
This project will require experience with JavaScript, familiarity with open-source JavaScript libraries for interactive maps (Mapbox GL, MapLibre), and experience with modern visualization libraries (D3.js). Experience with geospatial data formats is also desired.
The mentors for the project will include key staff from the City of Boston’s Office of Emerging Technology, including one who served as a Google Summer of Code mentor at Code for America in 2011 and the City of Boston between 2021 and 2024.
The City of Boston’s Parks Department maintains a comprehensive data set on trees in Boston. However, it’s a manual and laborious process to get the data (such as conducting site visits for tree counts) on a regular basis. Previously, a student built a Boston-specific model to identify trees from aerial imagery; it was based on Deep Forest, a machine learning library for tree crown detection.
This summer, we would like to operationalize this project for the Parks Department. Right now, the machine learning model runs on Google Colab, which has proven difficult for staff at the Parks Department to learn. They need a simple web interface that allows for the upload of updated imagery and subsequent analysis.
We would also like to explore ways to find more insights from the aerial imagery. For example, we are interested in the feasibility of determining tree health and the variety of tree species across the city.
The ideal outcome would be 1) the creation of a simple web interface for the Parks Department to upload new aerial imagery for analysis, and 2) the ability to the Parks Department to generate a list of statistics on tree counts and tree healthto ensure that it continues to plants trees in an equitable manner across the city.
We give this project a medium level of difficulty. The project can be completed in 175 hours.
This project will require intermediate experience with Python, machine learning (in particular TensorFlow and training models with imagery), interacting with RESTful APIs, and limited experience with JavaScript, HTML, and CSS.
The mentors for the project will include key staff from the City of Boston, including one who served as a Google Summer of Code mentor at Code for America in 2011 and the City of Boston between 2021 and 2024.
The City of Boston’s Transportation Department relies on labor-intensive manual processes for parking studies, where City planners walk Boston’s neighborhoods with clipboards, recording license plates at 15-minute intervals. This methodology limits the scope and depth of curb usage analysis, providing only a narrow snapshot of parking patterns in our urban areas.
Boston’s Office of Emerging Technology has identified an opportunity to change how the City conducts parking studies. By applying computer vision technology to analyze video feeds from standard cameras like GoPros, we can expand both the geographical scope and temporal range of curb usage analysis across Boston, while simultaneously reducing manual effort.
A Google Summer of Code contributor will prototype a tool that:
Future versions of the tool will:
The technical implementation will focus on evaluating computer vision APIs and/or training and optimizing object detection models (YOLOv8 or others) for urban curb usage analysis, with direct support from engineers and data scientists with the City of Boston’s Office of Emerging Technology and Department of Innovation and Technology.
We give this project a medium level of difficulty. The project can be completed in 175 hours.
This project requires experience with computer vision technologies, particularly in training and implementing object detection models. The contributor should be comfortable working with video processing pipelines and basic statistics.
The mentors for the project will include key staff from the City of Boston’s Office of Emerging Technology, including one who served as a Google Summer of Code mentor at Code for America in 2011 and the City of Boston between 2021 and 2024.