Sensibite

Inspiration

A few months back my close friend was almost on the brink of death due to a food allergy caused by foreign substances at a distant place where he wasn't aware of the food he was consuming, which led me to think couldn't he have been sensible before taking that bite?? That was the time when I came up with the idea of "Sensbite"- a food allergy assistant app that quickly lets users scan the food items they are about to consume whether it is packed or unpacked and tell them about the allergen contents of it.

What it does

Sensibite is an application that scans packaged and unpackaged food and checks for allergens that can cause allergies using a vast database of foods, the allergens they contain, and the problems they can cause. Case Study: Consider a case where you are traveling to a foreign location where you are completely new and unaware of the food you are planning to consume, you can't get the allergen contents of the food you are about to eat unless you google the whole thing which is not a preferred way and the language is a barrier between you and food seller to directly know anything about the food. In such a case, a quick application like Sensibite offers the best approach

How we built it

The application is built using Kotlin as the primary language for the Android platform with Firebase as a Backend service provider with the modern industry-standard MVVM architecture and powerful jetpack libraries for real-time barcode scanning and camera support. For the barcode scanning we are using open food facts API to get access to their massive database and to empower the most important feature of unpacked food scanning a trained Machine Learning model has been implemented to empower the process. The UI might not be the best but it definitely is User Friendly and and convenient by all means.

Challenges we ran into

The number of food items and dishes in this world is a large number which is hard to collect and needs to be updated from time to time, in order to serve the masses.

Accomplishments that we're proud of

With a huge amount of data, we have managed to get around the accuracy of 76% and there is no one in the market that is working on the allergens related to unpackaged food dishes allergen detection.

What we learned

During the fun and challenging 24-hour HTM 4.0 hackathon, we learned a great deal about the food industry and the essence of working in a team as it is our first hackathon participation and we came across various interesting technologies while getting our hands on this project.

What's next for Sensibite

Sensibite plans to further enrich its database and empower the marketplace feature for in-app purchase of healthy food items to make it a market-fit application

Built With

• camerax
• firebaseauthentication
• firebaserealtimedatabase
• jetpack
• kotlin
• mlvisionapi
• mvvm
• openfoodfactsapi
• tensorflowlite