During the COVID-19 pandemic and quarantine, physical activity in the world has gone down. During these times of isolation, it’s important for everyone to stay physically active due to the benefits that it offers: ease of muscle strain, relief of mental tension, improves blood circulation, and improves overall well-being. According to PhysioPedia, a renowned physiotherapy website, physical activity during COVID-19 is extremely important as it reduces the risk of several chronic diseases, such as cardiovascular disease and diabetes. As passionate students with an interest in maintaining physical health and well-being, we sought to create an application that helps people stay physically active. We believe that our platform can help people stay active and stay healthy during these unprecedented times!
What it does
This mobile app aims to reduce the burden on the medical industry during the COVID pandemic. It does this by offering exercises one can do for aches and pains. We believe this is a helpful app because at home it’s easy to stay in bed all day and not stretch, so pains are expected. Our physiotherapy platform, FlexClinic, helps people stay active and healthy during this quarantine. We provide various exercises, stretches, and routines to keep people fit and relaxed. Our machine learning algorithm detects key points on the user’s body and guides the user through exercise sessions.
How it does
The app collects data (symptoms) from the user to detect which parts of the body are sorest. From this information, an ML model picks a class of exercises for the user to complete. Then it uses the camera to trace key points on the body for several exercises targeting the user's injuries, step-by-step. After a hardcoded number of reps, it takes the user to the next exercise.
How we built it
We identified the strengths within our team and divided the work amongst ourselves to develop FlexClinic. We used Java for our back-end and Android Studio for both our front-end and back-end. In addition, we used Fritz.ai, and the MonkeyLearn API to build the machine learning and artificial intelligence components of our application.
Challenges we ran into
Throughout the development of our product, our main challenge was changing our technology and software stack to suit our requirements. We initially started using React.js, Tensorflow.js, and Bootstrap but we realized that several components and features we wanted to include were incompatible. After brainstorming and researching more about different technologies, we decided to use Android Studio and Fritz.ai as they worked best to implement our plan. Also, navigating between pages on the front-end of the Android app was tricky. The back-end was also difficult because we have little experience using Java so coding classes in the realm of machine learning and computer vision involved a lot of debugging.
Accomplishments that we're proud of
Our team is composed of a variety of students: 2 high school students from the United States and 3 university students from Canada. We are most proud of the collaboration we’ve had and we developed interdisciplinary solutions because we had people from different backgrounds.
What we learned
Practically speaking, while we were brainstorming ideas, we noticed the massive amount of benefits derived from physical activity but were stunned by the dip in rates of physical activity during the times of COVID-19. After doing some research about how people can be guided to stay physically fit, we found that an AI/ML approach of physiotherapy. Technically speaking, we developed our skills with Android Studio, React.js, Java, and using various ML repositories. We also learned a lot about augmented reality by experimenting with AR/VR technology as well.
What's next for FlexClinic
In the future, we would like to add more exercises for users and spread awareness about staying physically active in these isolated times. In addition, we would like to add a hospital/patient rerouting feature to our application to reduce patient traffic. Another idea we have is a clinical outreach program, which allows users to connect with licensed therapists and seek medical advice.