B23 Integration of multiple data sources for swimming biomechanics

Abstract

Assessment of swimming biomechanics is challenging because the activity is water based and is not stationary, nor can easily be simulated as such. This paper introduces sensors fusion tools to combine several technologies to enable a more robust investigation into the movements of swimmers. Swimming assessment is traditionally a labour intensive process where even the primary measures of stroke phase, stroke rates, stroke counts, and lap times have to be manually recorded or extracted from video data. While more detailed stroke information can be gathered from inertial sensor and video data it is time consuming and manually based. Recent advances in microsensors have demonstrated the validity of using wearable sensors to replace these manual methods in an automated way [1] and the potential to further extend the capability to detailed biomechanics measures [2] though the use of multiple sensors per swimmer. The use of microsensors however introduces new complexity through the collection of large amounts of data that need to be collected, saved, synchronized, and appropriately conditioned prior to analysis. Analysis techniques include filtering, event detection and display in a form useful for a range of audiences including athletes, coaches, sports scientists and researchers. This paper describes tools developed that facilitate the use of microsensor technology and its fusion with other data sources.