Speed skating technique and movement has been previously analysed, however, there is a notable absence of a real-time feedback system that provides athletes with quantitative insights into their performance. Therefore, this thesis aims to identify and measure the Key Performance Parameters (KPPs) and create a framework of a feedback system that allows coaches and athletes to receive real-time feedback during training sessions. To achieve it, IMU data of skaters were recorded, and algorithms were developed to quantify KPPs. Furthermore, an interview was conducted with a speed skater to determine the optimal presentation format for feedback. To validate the accuracy of the KPP detection algorithms, video data capturing skating bouts were used as a reference. The algorithm was able to detect the contact and air phases with a mean difference of .002 seconds, while paired samples t-test showed no significant differences in contact (p=.842) and air (p=.815) phases. Although the estimation of skate lean angle presented challenges, promising indications for its potential were observed. Through detailed investigation, the optimal mode, content, timing, and frequency of feedback delivery was identified. Consequently, this thesis has laid the foundation for an evolutionary tool for speed skating training and performance assessment. The combination of IMU technology with advanced algorithms and feedback mechanisms has the potential to revolutionise how athletes approach their training.
© Copyright 2024 Published by KTH Royal Institute of Technology. All rights reserved.