A multisensory datafusion-based 3D plank-coaching system

Exercises have become an important part of many people's daily life. However, inappropriate exercises cannot help exercisers build desired muscles, and may even get them hurt. In this paper, we proposed a 3D plank-coaching system to guide plank exercisers through fusing multimodal data and providing proper haptic feedback to them. Our proposed system can measure and fuse the user's muscle movements, posture, and several physiological parameters information, which are collected from Electromygraphy (EMG) sensors, Kinect v2 sensors, Electrocardiography (ECG) sensors, and thermometer. Based on comparison results between the user's fused data and the standard data, our system can activate the corresponding vibrotactile actuators, embedded in our developed haptic clothes, to suggest the correct adjustments. As a result, our system can improve the user' plank exercise quality and decrease his/her injury risk.