#The Saga of Arena Aligner
Our team covers a broad scope of expertise and interests; We wanted to pick a project where we could all shine in our areas of choice. Ryon and Toby are math majors who were able to expertly design an objective function to optimize our stadiums capacity and location. Kevius wants to drive the future of computer vision so he decided to apply computer vision to for land surveying. Wyatt enjoys playing around with large language models, so he was the creative liaison for our real estate agents storytelling ability. Drawn together by our love of sports; and desire to bring sports to regions without large teams. We decided to build a virtual real estate agent for games and recreation; that we believe could revitalize a smaller city.
Arena Aligner helps cities to grow in wealth and culture by identifying how they can expand into the world of sports. We have created a virtual real estate agent that proposes a optimal sporting real estate. Our agent starts by gathering data and identifying the optimal location in a city to build an NFL football stadium. Finally our agent tells a comprehensive story of how a new sports arena could enhance the city's culture by creating a team that reflects the city's unique spirit. It also will explore the financial insights of the stadium and how it could help bring in income for the city through various sporting and community events.
The core of Arena Aligner is the optimization function, which finds the best place in a city to place a new sports stadium by considering various factors, which are mainly extracted from aerial image and local municipality data. We developed this model by first brainstorming different factors that be optimized for in an objective function. As a final touch on the project, we used multiple generative AI models, including a vision language model, to provide details and insights about the value that building a stadium at the suggest location can bring to the city.
- There was a lack of nationwide geospatial data on zoning, landuse, and traffic flows. We could not utilize local data because we wanted our model to work on any city in the United States.
- Lack of data on buildings in cities: Solution segment and count number of buildings in a city.
- Lack of access to high quality satillite data: Solution take small samples using close to earth satallite imagery from NAIP
- NAIP is not Frequently updated and sometimes has lapses in data.
- Lack of easily accessible data
The amazing aspect of Arena Aligner is that it can be easily adapted to various different scenarios. For instance, it can be applied to any city in the United States or abroad. It can also be used to find spots for other kinds of venues, such as ones for different sports or even events such as concerts. Given more time and resources, we could enhance Arena Aligner by accounting for even more features, such as highway distance, allowing for a more optimal location suggestion.