Fitness Planner

Inspiration

Persistance is what matters the most when it comes to fitness. A good number of people fail to stick to their fitness plans because they occationally feel lazy. Hence they are unable to reach their fitness goals. I also suffer from such laziness, so I decided to fix this issue by creating an app that help people create a fitness plan and help them stick with it.

What it does

The main page of the app shows days of a week. When the user completes all the activities of that day, they can click done and put a check mark on that day. In this way, they can keep track of how many days they have been going to the gym. Our app only accepts check-ins when the user is within 200 meters of their selected gym.

How we built it

We used FlutterFlow to create the UI because it is surprisingly time-efficient. We also used Dart to code all of the beckend functions. Moreover, we integrated OpenAI API to generate the fitness plans and we used Google Maps API to pick the location of the users' gym.

Challenges we ran into

It was our first time creating an app and using FlutterFlow. We had to learn almost everything in a short amount of time. Integrating APIs OpenAI and Google were also challenging. It was also difficult to convert latitude and longitude differences to actual distances.

Accomplishments that we're proud of

We never gave up as we overcame all of the challenges we faced. The app works as intended and I planning to use it in the future.

What we learned

We learnt the basic workflow of creating a software, as well as using tools like FlutterFlow and Android Hub. We also learnt how powerful AI coding tools are and how to utilize them to speed up our work.

What's next for Fitness Planner

The next step for Fitness Planner would be its main feature, which is letting the AI coach to correct exercise motion. As a guy who hurt his back in his first year training, I understand that an extremely common mistake beginning body builders shares is that they often fail to keep the quality of their training sets, which makes them prone to hurting themselves. Thus, I want to build an AI tool to supervise them. A rough thought of this direction is to convert human postures into graph, and use language model to distinguish bad motions to standard ones, which is essentially explain the difference of complex data structures.