Many pets are not given a proper diet because it is hard to measure the right amount of nutrition that is consumed by the pets everyday. Most pet owners decide to not have regular checkups or not have veterinary visits at all, regardless of how important they are. These compromises result in health conditions of the animals involving digestive upset which in turn results in issues such as severe obesity, diabetes, pancreatitis, and more.

What it does

PetCare strives to alleviate this issue by providing a cross-platform mobile application to pet owners that can help them maintain their pet's health accurately and efficiently. With modern technologies, our app achieves this through two methods: first is recognizing if a pet food item is healthy given the pet's current health and second is monitoring the pet's health with parameters such as protein consumed, amount of fat and fiber consumed, and much more. With this service, it is expected that a large number of pet owners will take advantage of the modern technology that this application utilizes in order to keep their pets healthy without regularly being dependent on professionals for guidance. Even though experts' guidance can be effective, this app provides an alternative to constantly track the pet's health and guide the users to make healthy choices for their pet, removing the necessity of frequent visits to professionals.

How we built it

We build our cross-platform application using Flutter, a framework by Google, along with firebase and other dependencies. We used Firebase Authentication for using Email/password login along with google login to authenticate users. Firestore was used to save, read, and real-time update the user's information. To display their information, the flutter charts dependency was used which enabled us to create beautiful and functional charts within the app. The main feature of the app is a product evaluation. This uses Firebase MLVision that can read text from an image using an OCR machine learning model. It then uses this information and goes through a filtration algorithm to remove excess data. The important data then goes through the custom algorithm (proven effective by research) to decide whether the user should buy the product, given the pet's current health. This way, their pet's health is kept consistent or improved over time.

Challenges we ran into

The main challenges we ran into were related to the Firebase MLVision. Cocoapods gave us a hard time working on the app since they kept causing bugs that we had to fix. One fix made and another 10 bugs were made. We were able to get around it by learning the pattern and understanding what is causing the bugs, in order to eventually solve and finish the app.

Accomplishments that we're proud of

We are proud to create a platform that currently does not exist in the market and has a lot of potential to improve the lives of many people. Using this app, people will not have to worry about frequent veterinary visits for their pets in order to maintain proper nutrition, since PetCare is able to do exactly this. We were able to, in the end, simplify the lives of pet owners who now have fewer responsibilities to take care of.

What we learned

We learned many things about how the machine learning model is able to recognize text using deep learning. We were able to understand why very few pet owners have veterinary visits and what demotivates them from going. Taking the advantage of this, as we built the app, we also learned more about the structure of the framework itself, since solving the bugs really caused us to dive deeper not only in the framework but how iOS app development works individually as well.

What's next for PetCare

In the future, we would like to publish this app for public use. As time goes on, we would like to add more premium features that improve the quality of the app and functionalities that make it easier to do things currently harder to do.

Built With

Share this project: