Inspiration
Housing and housing related costs is one of the largest expense categories for people. Buying furniture is also a big hassle: visualizing a great design as you walk through a store is difficult, and deciding if the furniture will fit in your house can be even harder! We can improve the customer experience, give ideas on how to decorate, and cut down on boxes, travel, and returns with technology.
What it does
We give people the ability to place items around their house from a database of household objects/purchases. They can scroll through suggested room setups, and instantly see how the furniture fits in with what they have using AR. This helps reduces emissions for shipping and trips customers make for purchasing and returns. We can also point customers to specific stores and help them find good deals.
How we built it
We designed and developed a mobile application that employs AR to help users create the home of their dreams! We used Flutter/Dart and ARCore for the front end, FireBase for our database solution, and a python Flask server for our backend.
Challenges we ran into
We had many challenges. We intentionally chose technologies that we were unfamiliar with in order to learn. Setting up and integrating the projects was especially difficult, as we had to come up with the architecture, improve our lack of knowledge, and deal with a lack of sleep at the same time!
Accomplishments that we're proud of
Majority of us had a limited amount of experience in all of these fields. That made it a big accomplishment any time anything worked. Our project has a very visual payoff, so seeing the furniture and home decor displayed right in front of us was extremely rewarding. We were able to successfully train a Linear Regression learning model, and predict the best customized floor-plan for a user.
What we learned
We learned ARCore, flutter, how to connect the two together, how to link with Firebase, and how to include a python backend.
What's next for FurnishAR
We really hope to improve our UI to present to users. We hope to also increase the database of products we offer customers and potentially integrate with real stores. In terms of machine learning, we hope to gather more data from our customers so we can create a multi-layer learning model, which would better personalize the customer experience!
Log in or sign up for Devpost to join the conversation.