Boilerfit

Boilerfit—designed by a Co-Rec trainer, for Purdue students. Get personalized workouts using Co-Rec-specific exercises, smart tracking, and injury prevention. Train smarter and maximize gains today!

Comment

Inspiration

The inspiration for Boilerfit started during my senior year of high school. By that point I have been training in powerlifting for years, broken 6 Pennsylvania state powerlifting records, and was coaching other clients/friends on powerlifting. I started bringing my friends to lift with me more. I started realizing that I really love training other people and that if I became a personal trainer at Purdue as a side job that it would be very fun and fulfilling for me. So I started studying for my personal training certification in the summer and eventually became a personal trainer at the Purdue Co-Rec. Personal training is an incredible job and i'm really grateful for it, but I recognized that the price of personal training is not for everybody and that information regarding lifting and how to program everything can be wildly misleading depending on who you listen to. Because of that, the idea for Boilerfit was born. Me and my teammate and friend Jack Smith created this application that is essentially an autoregulating fitness app designed just for the Purdue gym as every single exercises that can be done at Purdue's gym are the exercises allowed within the application. The app has settings on experience, training style, low/medium/high volume training, types of split you want, and all of those settings help you generate a split that you will follow for your training. And depending on how you do within regards of weight, reps, and intensity, there's a linear regression model behind it that tracks your progression and creates an expected weight, reps, and intensity for the next time that you do this workout. And that if you fall too short behind the expectation, it implements a deload for recovery as the model would've detected overtraining. The good thing about the model is that the more data you put in, the better it gets at detections. Within each of the exercises listed for the workout, it has the training type (Strength, hypertrophy, both), compound or isolation type, difficulty (beginner, intermediate, advanced), rating (1-3 stars), description of what the exercise does and how to perform it, image showing how to do the exercise, and a map of the Purdue basement gym that lights up the part of the gym that you are able to do that exercise at.

What I learned

The biggest thing I learned while going through the project is just how many specific details that you have to take into consideration when generating a split for someone. Things like experience, training style, volume per set, volume of sets per week per muscle, days per week going to the gym, etc. I knew that from experience of building program for my client that there are a lot of factors to go into consideration, but not at the level that Boilerfit takes into consideration.

How I built the project

The project was built using flask, python, sqlite, javascript, tailwind, HTML, and CSS. The SQlite held the data for all possible exercises and the data for it, which was found by me mapping out every single machine, free weights, barbells, and equipments in the basement gym prior to the start of the hacking. For the backend, Flask was used to handle API requests and process workout data, while SQLite stored all the exercises and user progress data. The frontend was built with HTML, Tailwind CSS, and JavaScript to provide a clean and responsive user experience. The application features an algorithm that generates personalized workout plans based on user preferences, experience level, and training goals. Additionally, the linear regression model helps adjust training intensity over time by analyzing user performance data. Throughout the development process, I learned a lot about integrating machine learning concepts into a fitness app, structuring a database for exercise-related data, and designing an intuitive user interface for ease of use.

Challenges faced

The challenges that were the most common throughout building the project was making sure not to make tiny mistake when processing the workout data as just the raw amount of exercises and all of the data that it contained made it easy to make small mistakes that ends up hurting in the future.

Future status of the project

Me and my teammate Jack fully believes in the product and the idea of a purdue-exclusive gym app. We will be working on this project throughout the rest of the semester and summer to hopefully deploy a fully completed project that we are 100% confident to be able to satisfy all customers in regards of working out at the Co-Rec with the assistance of Boilerfit.

Share this project:

Updates