Online Activity Matching Using Wireless Sensor Nodes

In this paper, we explore the capability of wireless sensor networks to perform online activity matching for sport coaching applications. The goal is to design an algorithm to match movements of a trainee and a trainer online and to find their spatial and temporal differences. Such an algorithm can aid the trainer to better observe performance of the trainees in group lessons. We consider fitness-like movements such as those performed in aerobic. We also limit ourselves to only having one sensor node on the trainer and one sensor node on the trainee, however our algorithm scales well to more trainees per trainer. We use Sun SPOT sensor nodes and use the accelerometer and gyroscope sensors to capture the movements. The gravity vector is extracted and improved with a Kalman filter using the accelerometer and gyroscope data. An automatic segmentation technique is developed that examines the movement data for rest and activity periods and changes in movement direction. The segmentation and the movement information are communicated with the node of the trainee where the movements are compared. We choose to use Dynamic TimeWarping (DTW) to perform the spatial and temporal matching of movements. Because DTW is computationally intensive, we develop an optimized technique and provide feedback to the trainee. We test all the design choices extensively using experiments and perform a system test using different test methods to validate our approach.