As a group, we felt that people with dietary restrictions always have had a difficult time finding food that matched their religious or personal beliefs. eatSafe sets out to solve this goal by integrating data from many of the world's most popular and trusted nutritional databases to create a single, streamlined, cross-platform application that allows users of all backgrounds, regardless of culture, religion, or ethnicity, to easily identify foods that align with their preferences.

What it does

eatSafe is a cross-platform mobile application that helps users find food that fits their dietary restrictions. This is done by combining the power of Google Cloud's ML Kit along with powerful REST APIs that communicate with the world's largest database of nutrition and food. When first entering the application, users enter in their dietary preferences, which are split between allergies and lifestyle choices (religious or personal). From there, users are presented with a camera-view, which they can point to the barcode that is included on all FDA-approved foods in the United States. After scanning the barcode, users are presented with a quick decision on whether a food aligns with their preferences, as well as the reason why, with an ingredient list that is included.

How we built it

Our team began the process of creating this application by whiteboarding the design of the application, as well as ideating various features in the solution. Next, we distributed tasks amongst ourselves, such as wireframing using Figma, initializing the project and its dependencies using React Native (built for iOS and Android), and setting up the backend infrastructure using JavaScript and Firebase MLKit. Finally, we tested our program on a variety of devices and foods to test for edge cases as well as ensure the reliability of its basic functionality.

Challenges we ran into

The biggest challenge we faced as a group was the clear lack of technological advancement in image recognition software, specifically in the use of barcodes. There were many instances where lighting and uneven surfaces interfered with the program's intended functionality. To address these challenges, we made sure our code handled these edge cases by creating engaging user interfaces that helped users better use eatSafe.

Accomplishments that we're proud of

Initially, we found the project to be overwhelming due to the number of new technologies that we would have to learn. However, through dedication and a commitment to gaining new knowledge, we are proud of the product we have released and are excited to iterate this program in the future using these new technologies.

What we learned

Coming into this project, we were unfamiliar with the machine learning models that were required to construct this product. Specifically, Firebases' image recognition model was new and untraversed for all of us. By the end of the projects, we are much more confident in our ability to implement machine learning models on a broad spectrum of use cases.

What's next for eatSafe

We hope to expand the number of restrictions that users may assign to themselves on signup. By doing so, we hope to achieve our mission of being more inclusive of wider and more diverse cultures.

Share this project: