Gyrofit

Interactive Fitness & Nutrition Game

GyroFit is a mobile fitness game that transforms your real-life movements into gameplay through gyroscope tracking. Designed for PC and mobile, the player uses their smartphone as a motion sensor to perform exercises while watching their avatar respond in real-time on screen.

The game begins by collecting basic user data like age, weight, height, and daily steps. This information is used by a machine learning model (trained with TensorFlow and deployed with ONNX) to generate a personalised daily workout plan — for example: “25 squats, 30 jumping jacks.”

Inside the virtual gym, players complete exercises such as squats, sit-ups, jumping jacks, and push-ups. Their avatar mirrors each move, providing real-time feedback. Points are earned based on accuracy and speed, encouraging users to improve their form and consistency.

Beyond exercise, GyroFit includes a nutrition section where users can calculate their daily calorie needs, access healthy recipes, and complete interactive awareness tests about food and health.

There are four different playable avatars (varying by age, gender, and body type), and a time-trial challenge mode to test speed and endurance. Weight change is visually tracked on the avatar model based on the user’s progress.

All of this is wrapped in a playful, stylised 3D world designed using Blender, Photoshop, and Substance Painter — making GyroFit not only functional, but fun

Grade: 85 (Presented as MSc Game Development Dissertation Prototype)

Technologies: