Tac-Trainer: A Visual Analytics System for IoT-based Racket Sports Training

Area of Science:

• Sports Science
• Human-Computer Interaction
• Internet of Things (IoT)

Background:

• Traditional racket sports training often suffers from subjective biases inherent in coach experience.
• Smart wearable devices leveraging IoT technology offer potential for data-driven training but face challenges with data volume and dimensionality.
• Existing methods for extracting information from sensor data lack actionable insights for coaches.

Purpose of the Study:

• To propose an integrated framework, Tac-Trainer, combining IoT data and visual analytics (VA) for effective racket sports training.
• To address the limitations of large data volumes and high dimensions in sensor data for practical coaching applications.
• To facilitate data-driven decision-making by coaches through actionable insights derived from kinematic data.

Main Methods:

• Development of the Tac-Trainer framework with four key components: device configuration, data interpretation, training optimization, and result visualization.
• Collection of trainee kinematic data using IoT-enabled wearable devices.
• Transformation of raw sensor data into meaningful attributes and indicators for analysis and suggestion generation.

Main Results:

• The Tac-Trainer framework successfully integrates IoT sensor data with VA to provide actionable training suggestions.
• The system transforms complex kinematic data into understandable indicators, aiding coaches in optimizing training.
• An interactive visualization interface allows for exploration of training data and results.

Conclusions:

• The proposed IoT + VA framework, Tac-Trainer, offers a novel solution for data-driven racket sports training.
• This approach mitigates biases in conventional coaching by providing objective, data-informed guidance.
• Future research directions include exploring VA for IoT data and IoT for VA applications in sports training.