Development of devices and applications used to identify the strokes of players

The institute IICT of HEIG-VD has gained a lot of experience on wearable sensors, like inertial measurement units (IMUs) that integrate accelerometers, gyroscopes, and sometimes magnetometers. 

These IMUs are typically worn around the wrist, but if needed, other units can be placed on a pocket, on a belt, on the chest, shoes, etc. By wirelessly acquiring the sensor data, a computer or a smartphone can characterize body movements, for instance the ones of a tennis player, by means of machine learning algorithms. 

The research at IICT on this technology started with the aim of supporting the quality of life of our increasingly older population, by developing an activity recognition system, intended as a compensatory tool for elderly people with memory impairment. The system acquires data from multiple sensor units, it recognizes various activities of daily living and provides a “browsable” log or history of activities on a tablet device. Similarly, the IICT team has applied the same technology and similar machine learning algorithms to diverse sport activities, including racquet sports and running. 

On the one hand, they developed an application that can identify the strokes performed by the player and compare the player's strokes between them or with a reference one. This can help players enhance their style and become aware of the most frequent strokes they perform during a match. 

On the other hand, they developed a prototype of a "personalized coach" application capable of providing real-time feedback to a runner based on his/her historical performance. These applications have been originally developed having in mind amateur sportspersons, but can be tuned for professional athletes.