Measuring relay exchange kinematics in short-track speed skating using a multi-camera network

To support their targeted improvement of the relay exchange, Great Britain Speed Skating required a tool that could be used to advance knowledge on ‘how to execute the relay exchange effectively’. A tool that measures relay exchange kinematics in representative race scenarios, over its entirety, and with an acceptable level of measurement error (± 0.19 m·s-1). A review of existing measurement solutions found that the Olympic Oval (CAN) multi-camera network was the only tool that came close to meeting this criterion. However, while this multi-camera network satisfied the metrics, scenarios, and scope of relay exchange measurement, its ± 1.53 m·s-1 error exceeded the target measurement error. For these reasons, this thesis developed a multicamera network to measure accurate, two-dimensional, relay exchange kinematics. The literature review identified that the accuracy of the National Ice Centre (GBR) multi-camera network was dependent on five sources of measurement error. Accordingly, a series of investigations quantified how these errors propagated, independently, to errors in relay exchange kinematics. In the case where these errors exceeded the target measurement error, additional studies investigated minimising each error. Using this empirically informed measurement workflow, Monte Carlo simulations showed that the multi-camera network’s total error was ± 0.17 m·s-1. This error was within the target measurement error and significantly less than the benchmark Olympic Oval (CAN) multi-camera network. Investigations into the execution of the relay exchange demonstrated how this reduction in error allowed Great Britain Short-Track Speed Skating to advance knowledge on ‘how to execute the relay exchange effectively’. In turn, supporting the team’s targeted improvement of the relay exchange, and ultimately, their aim of delivering medal-winning performances at the Winter Olympic Games.