Integrated Automation System for Speed Skating, Event Monitoring and Fault Detection

Abstract

Science and technology have taken the world to the next level and the need for automation in every field has increased equally. Thus, to make regional competitions of speed skating transparent and accurate, an idea to develop an automated system was discussed at Seinajöen Urheilijat, a speed skating club in Seinajoki. Consequently, this research was conducted to develop an integrated automation system that functions in accordance with the International Skating Union’s protocol and detects false starts, premature movement, delays and estimate the posture stability. The designed system uses a high-speed Sony FS camera and Python based frameworks. OpenCV and MediaPipe have been used for video acquisition and pose estimation respectively, whereas Pygame manages audio synchronization and NumPy checks the movement. A finite state machine is used in the system to manage all state transitions in the starting sequence of speed skating competitions, beginning with the audio command “Go to the start” and ending with either “Gunfire”(if there are no false starts) or “False Start”. The prototype has been able to achieve over 95% accuracy under indoor conditions and has successfully detected all fault starts and line violations and premature movement. The developed system is cost effective, lightweight and allows for a future expansion. Timing sensors and cloud-based analytics can be added to the system to make it more accurate and applicable.

The manufactured automation system significantly reduces human error, increases accuracy and transparency and will likely help the speed skating skaters in the club.