P-03

For most sports, injuries are often the result of overstraining. An important question is how these injuries can be prevented. This project aims to use sensor technology and big data analysis for the early detection of signals of overstraining in order to prevent injuries. To do so, the Saxion research groups Ambient Intelligence and Fitness & Health collaborate with Hanze University of Applied Sciences Groningen, Roessingh R&D, professional sports clubs (FC Twente, FC Groningen, Eurosped TVT and others) and partners working on sensor technology (360SI, CE-Mate).

Already, many technologies are used for measuring athletes, as a form of quantified self. Professional clubs invest in expensive systems to track and measure teams to improve their performance and prevent injuries. However, two main problems remain: first, the large amount of data and second, the knowledge that is required to interpret the data and convert these to a training advice. These problems can be solved by using computer models that are generated from systematic data analysis based on the gathered big data, combined with domain knowledge. This calls for a system that stores information from different sources and makes this accessible, so that the data can be combined, aggregated and analyzed. The system must build individual player profiles from these data for fast, automatic interpretation. These profiles can be used to estimate the likelihood of overstraining, so that trainers can adapt the training curriculum to prevent injuries.

Until now, the project has focused on measuring and predicting an athlete's rate of perceived exertion. Due to a low amount of actual injuries, the prediction of injuries has not been successful so far. For this reason, the project now focuses on internal player load. Building on domain expertise, we are creating models which will assist trainers to determine whether the load of a field training matches the experienced load of players.