RunTrack: A Real-Time Running Performance Optimization System using Temporal Convolutional Networks
Year: 2025
Type: article
Abstract: — RunTrack is a cutting-edge fitness tracking app developed using modern Android technologies, including Kotlin, Jetpack Compose, Room Database, and TensorFlow Lite. It transforms running performance analysis through a powerful machine learning backend based on a Temporal Convolutional Network (TCN) for real-time pace prediction. Room Database handles fast and reliable local data storage, while the Jetpack Compose interface delivers a responsive and elegant Material Design 3 experience for tracking activities, monitoring performance, and viewing data. Core features include live GPS tracking, intelligent TCN-based pace forecasting, detailed activity logging, performance metrics, and interactive charts with predictive insights. RunTrack is built specifically for mobile users and makes smart use of on-device machine learning, offering runners a practical and intelligent way to improve their performance and stay on top of their fitness goals. One standout feature is PacePredict, which uses real-time sensor input along with past running data to give runners pace suggestions tailored to their current condition. By applying advanced Temporal Convolutional Network (TCN) models, it helps identify signs of fatigue and adjusts pacing advice based on changing conditions. This smart blend of live tracking and predictive insights offers runners a well-rounded tool to enhance their training. With its user-friendly design and ability to scale and adapt, RunTrack makes it easy for runners to monitor progress and make data-informed decisions to improve over time. Index Terms— GPS Tracking, Real-Time Monitoring, Pace Prediction, Performance Metrics, Predictive Modeling, Machine Learning, Temporal Convolutional Network (TCN), TensorFlow Lite, Mobile Computing, Fitness Insights, Health Metrics, Running Data, User Experience, Secure Login
DOI: 10.55041/ijsrem47992
Authors: author, Dr. Nagaraj S R
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4410400302