OSUTrade

This project aims to create a website for Oregon State University (OSU) students to trade used items.

Features

Frontend

• Login required to view the full site
  
• Display of item pictures, product names, and prices

Backend

• Login using OSU accounts
  
• Upload pictures → generate 3D models

Inspiration

As an international student at OSU, I often trade used items with other students. During these times of inflation, it’s essential to find affordable items. However, most trading currently happens through social media groups, which isn’t ideal. My frustrating experiences trying to find items and compare prices inspired me to create this project. The idea used items trading platform must be trustworthy and easy to review. Thus, I plan to create a website that can only log in and review for college members and can use AI to generate 3D models with product pictures. Customers can trust these products from other students and review the details of the products by 3D models.

What It Does

OSUTrade is a secondhand trading platform for OSU students to buy and sell used items. The website allows students to upload photos of their items, and others can browse and purchase them.

• Calls backend APIs to display all products on the homepage
  
• Product detail pages include a Google Map (for the trade location) and a 3D model of the item

How We Built It

This is a full-stack application built with Next.js. The backend follows the principles of clean architecture. We have implemented the interface adapter layer (controller, repository), use case layer (service), and entity layer (model) to ensure the backend is easy to test and scalable for future enhancements.
The frontend uses the react-three package to load 3D models.

Challenges We Ran Into

We were limited by time and budget. A better computing environment and more time are needed to run the 3D AI model (TRELLIS) in the backend.

Accomplishments We're Proud Of

While it may not be immediately visible in a demo, using clean architecture allows us to easily add features and write tests in the future.
We also successfully created a website that can load 3D models quickly—something I’m really proud of.

What We Learned

• Gained deeper familiarity with the Next.js framework and clean architecture
  
• Explored the potential of AI models in computer graphics

What's Next for OSUTrade

• Continue building and completing the remaining pages of the website
  
• Deploy open-source AI 3D models in a cloud environment, allowing users to generate live 3D models from uploaded images
  
• Implement login/sign-up functionality to restrict access to OSU community members only
  
• Publish the MVP by this summer

Built With

• 3d-model
• google-maps
• next.js