Penguin Warriors

Inspiration

As busy university students who sit for hours at a time, our team noticed that we all suffered from lower back pain. Often, incorporating some quick, simple exercises into a daily routine can be the difference between feeling horrible and feeling great. This fact, along with our interest in fitness and wellbeing inspired us to create a desktop mini-game that reminds you to take intentional breaks from the screen to keep active.

What it does

When the game starts, the game prompts the player to complete a set of exercises for a specific duration. After completing these exercises, the player gains enough points to move on to the next stage. Moving on to the next stage would involve defeating different enemies or reaching a certain amount of points.

How we built it

Penguin Warriors uses the machine learning libraries TensorFlow and MoveNet to track user movements. Our team split into two focus groups, two members worked on the UI, while the others worked on implementation of the machine learning tracking. The UI was implemented using HTML, CSS and JavaScript.

Challenges we ran into

Initially, we wanted to work on a different idea that used a Google Maps API; however, we struggled to make progress with this topic. We realized we had to shift gears in order to create a tangible product. This led to us working with TensorFlow and MoveNet. However, these materials were still new to us so the learning curve was steep. We had trouble downloading and implementing these materials and came across many bugs. We solved these problems by asking mentors, relatives and following YouTube videos.

Accomplishments that we're proud of

We are proud that we managed to make a working product despite changing ideas halfway through the hackathon. We are also proud that we were able to learn about many different APIs and about machine learning motion tracking. As the writer of this post I am proud of my teammates for demonstrating perseverance, grit, and cleverness in the face of immense difficulty and stagnation. Despite setbacks, our team kept our spirits high and made it through mostly in once piece.

What we learned

Besides exploring TensorFlow and MoveNet, we learned how to optimize our planning strategies. For future hackathons we learned that we need to establish our strengths, experience level and interests initially, so that we can generate unique ideas that cater to our skills. We also learned that we need to focus on modularity so that we can minimize the overlapping of each person's work. This is important because we often had merge conflicts that were confusing to solve.

What's next for Penguin Warriors

Hopefully in the future we can develop Penguin Warriors to be a robust application with customizable exercises, time durations and higher performance. It would be wonderful to continue development such that our application is compatible with both phones and computers. Additionally, we would like to add additional features such as maintaining streaks, and reminders to take breaks from sitting.