Interaction design for automatic tracking of free weight exercises using computer vision and IoT systems

Sammanfattning: Computer vision is on the cusp of revolutionizing several different markets through its applications. One such market is that of fitness and exercise. Advagym, Sony's connected gym solution is currently employing IoT solutions to track exercises in a gym environment. Together with Sony's R\&D Center Lund Laboratory's computer vision- and machine learning-equipped tracking system - they are exploring solutions to free weight exercise tracking. Using a human-centered design process - several iterations of prototypes with varying degrees of fidelity were developed, in order to find and evaluate a solution that would enable Advagym to offer an automatic free weight exercise tracking solution to its users. The final iteration of the prototypes was implemented as a fully-functional smartphone application running on the iOS and Android operating systems. The application leveraged both Advagym's IoT devices and the Sony's R\&D Center Lund Laboratory's tracking system in order to deliver on its purpose. Four final prototypes were put through a usability test with 20 participants in order to evaluate several modes of interaction and user interfaces that enabled the automatic free weight exercise tracking. The test showed that the prototype had an above average user experience - and 90\% of the test participants would like to use a system such as the one proposed, if it existed on the market.