Inspiration
Fast fashion is an extremely significant contributor to climate change. Currently, it is responsible for about 10% of all carbon emissions. The average American discards over 65 pounds of clothing per year, so there is clearly plenty of clothing already in circulation. We decided to combat the fast fashion industry using technology by providing an easy way for consumers to identify secondhand alternatives to their desired products. After brainstorming and discussing, our team envisioned a website that would streamline the process of discovering secondhand clothing based on any fast fashion item a user encounters.
What it does
GreenThreads enables online shoppers to access alternative clothing options from secondhand markets, leading to reduced overall carbon emissions. Our website gathers user input, which we then filter and display eco-friendly variants from pre-owned vendors including Depop, Mercari, and ThredUp. Not only does the consumer help improve the environment, but they also save themselves money and support individual sellers/small businesses. We aim to continue to promote sustainable living and reduce fast fashion as a whole.
How we built it
Our project uses several Python libraries/frameworks (flask, selenium, BeautifulSoup), REST APIs (serpapi), React libraries (TanStack Query), and BootStrap for the UI/UX components. Our team split into two divisions for the frontend and backend which streamlined development. The dataflow of our full-stack application begins with the user inputting a link, which is passed from the front end through a POST request to the backend. The backend scrapes data from the website, finds alternative clothing options, filters unsustainable brands, then returns the list of products back to the frontend. We used Git for version control and Figma for UI design.
Challenges we ran into
The biggest challenge we encountered was properly scraping data from the user-inputted website. The description of each product is not consistent across all websites, so we were forced to search for solutions that would yield the most accurate output no matter the layout of the site. One attribute that was especially hard to identify was the color of the item. We compiled a list of common colors and searched the visible text for matches, then returned the result closest to the top of the page. The second problem we ran into was integrating our frontend and backend. It was difficult for us to make the program run on everyone’s systems because of our varying environments, but we eventually managed with some tweaks to the code.
Accomplishments that we're proud of
Our website appeals to a broad audience, whether their motivation is to save money, pursue sustainability, or simply a preference for second-hand clothing. We are especially proud of the UI of the website and how we made it function smoothly with the backend. We are glad to deliver an end product that is highly functional and visually appealing in only 24 hours.
What we learned
Our team learned to work cohesively under a tight deadline. We found that compartmentalizing the project into many small problems made it easier to stay organized. Communication was especially important due to the amount of people working on the same program. There were a few mishaps with git branch management which were quickly resolved with clear communication. We also learned a lot about the processes involved in web scraping and how to effectively code workarounds when dealing with unreliable data.
What's next for GreenThreads
Our next step is to expand the tool into a Chrome Extension and find alternatives as the user shops. We also plan to integrate image searching into our algorithm to improve the search results as well as develop an algorithm to quantify savings in carbon emissions. In the future, we believe that we can take GreenThreads from a hackathon project to a fully fleshed-out tool for sustainability.
Log in or sign up for Devpost to join the conversation.