Spotly

Inspiration

Parking in NYC is inefficient, time-consuming, and environmentally costly. Drivers spend valuable time circling blocks, contributing to congestion and increased carbon emissions. We were inspired by the idea that open city data and predictive analytics could transform this everyday frustration into a smarter, data-driven urban solution. As students passionate about technology and data, we saw an opportunity to improve mobility and sustainability in one of the most complex cities in the world.

What it does

NYC Parking Predictor uses predictive analytics and public datasets to forecast parking availability by neighborhood and time of day. Instead of guessing or circling endlessly, drivers can view high-probability parking zones before arriving at their destination. The platform reduces search time, traffic congestion, and emissions by guiding users to smarter parking decisions.

How we built it

We used NYC Open Data datasets (such as parking violations, street cleaning schedules, and traffic patterns) to identify trends in parking availability. After cleaning and preprocessing the data, we trained a machine learning model to predict parking likelihood based on time, location, and historical behavior. The frontend displays results through an interactive, user-friendly map interface, allowing drivers to visualize predicted availability in real time.

Challenges we ran into

One major challenge was working with incomplete or indirect datasets, since NYC does not provide real-time parking availability data. We had to creatively infer availability using proxy indicators like violation frequency and street regulations. Data cleaning and feature engineering were also time-intensive due to inconsistent formatting and large dataset sizes.

Accomplishments that we're proud of

We successfully built a working predictive model using real public data and translated complex analytics into a simple, intuitive interface. We’re especially proud that our project not only demonstrates technical skills in machine learning and data engineering but also addresses a real civic issue with measurable environmental impact.

What we learned

We learned how to work with messy, large-scale public datasets and transform them into actionable insights. We also gained experience in feature selection, model evaluation, and translating technical outputs into user-centered design. Most importantly, we learned how data can directly improve urban life when applied thoughtfully.

What's next for NYC Parking Predictor

Next, we plan to integrate real-time data sources such as IoT curb sensors or crowdsourced availability updates to improve prediction accuracy. We also aim to optimize the model using more advanced algorithms and expand the platform into a mobile application. Long term, we envision partnering with city agencies to support smarter curb management and sustainable urban planning across NYC.

Built With

• claude
• streamlit