Basketball Shoot Tracking App with BioMechanical Device Integration
A comprehensive tool to track and improve ball shooting form with real-time AI analytics. The project included integrating a Meta MotionS device to capture data and display them on a separate companion application for detailed analysis and shoot improvement tactics.
A company that utilizes AI technology to enhance basketball training
Enhancements
Arm Motion Insights
Shoot Tracking
Coach Assistance
Services
Sensor Integration
Angle-Based Algorithm
LSTM Neural Network
Tech Stack
Swift
iOS Native
Python for ML
Project Overview:
Perfecting Free Throws
Free throws and shoot accuracy are crucial elements that identify how good a basketball player is—the proper biomechanics involved in two types of shots; overhead push shot and underhand loop shot while training holds more superiority over an official game.
IoT devices track precise biomechanics and motion metrics, while AI uses this information to refine techniques. Folio3 Digital Health specializes in seamlessly integrating IoT wearables with custom AI models to deliver intuitive, real-time feedback through scalable, secure mobile apps.
Training is where players overcome all their mistakes and upskill, while the official game gives them a platform to shine, which is only possible if they have mastered their basket-scoring options and tactical dodges.
Folio3 helped BLLR make such tactical and technical training possible, allowing their end consumers, “Basketball Players” and “Coaches” to not only track their full motion data along with their shot sequences but also get involved in perfecting their shots and feed it to their subconsciousness so their natural game becomes their best game when playing under a lot of pressure and jittery official games.
BLLR sought a tailored solution to elevate basketball game performance by capturing and studying players’ arm movements during their attempted shots.
Traditional methods lacked precision, relying on visual observation and video analysis. Our companion application, integrated with MetaMotionS, captures motion data and provides insights to improve shooting angle, speed, and projectile motion. They faced issues in:
Sensor Integration: We synced MetaMotionS with an iOS app to collect motion data from the magnetometer, gyroscope, and accelerometer. Later recording free throws with higher precision. Result:Players gained precise motion data that served as the foundation for performance improvement.
Angle-Based Algorithm: An algorithm was developed that uses the arm angle to point out the start and stop points of each throw, enabling players to compare their shots against an ideal benchmark. Result:Players could immediately identify key differences between their shots and an ideal form, improving shot accuracy over time.
Advanced Machine Learning Model (LSTM): An LSTM neural network was used to detect shots within continuous basketball motion data, capturing the complete shot sequence for more comprehensive comparisons with ideal shots, enabling players to refine shooting consistency and technique.
Data Collection & Training Process: Two data types were used: primary (training) and secondary. Primary data consisted of 20 free throw shots from two players. Secondary data, collected from two minutes of dribbling, trained the model to differentiate between shot and non-shot motion.
Enhanced App Capabilities: The app enables players to explore trends in their shooting form and develop their peak performance over time. Moreover, allowing coaches to access detailed data on each player’s performance to help them highlight key improvement areas.
Optimized Performance: Players with the help of data-backed insights into their unique motion sequences, witnessed a significant improvement in their shooting accuracy.
Injury Prevention: By cumulating the redundant and atypical motion patterns, the app highlighted potential injury risks, providing coaches with the insights to tweak their training and refrain from strain-related injuries.
Potential for Future Expansion: The app’s modular architecture allows for further enhancements, including comprehensive full-body motion tracking, analysis of various other basketball shot types, and its implementation across other sporting activities.
