Digital Sports Bavaria
Abstract
Digitalization becomes more and more part of different areas in daily life, e.g. sports and fitness: Wearables, nowadays especially step counters or activity trackers, support humans by individual promotion of health and improvement of fitness. Due to the various application areas, these Wearables are a profitable market for the future, but current existing systems have the following disadvantages:
Currently available sensors on the market are external devices which are manually put into apparel. An incorrect use, based on wrong placement of the system, leads e.g. to movement patterns of steps which deviate from a usual step. This deviation is a challenge for further processing.
Signal analysis is mainly based on the classification of movements, e.g. the detection of steps. In many applications, the quality of movement is important. Based on the information about, how a step is performed, an assessment of e.g. the state of health of a person can be performed.
The user interfaces are mainly based on simple text and/or graphical output on mobile devices or PC, e.g. number of steps per day and during one week. Nevertheless, for long-term success of systems, innovative approaches are needed.
Wearables mainly communicate with mobile devices, e.g. smartphones or a PC. Data are transmitted to a server via web browser for storage and display. This kind of system architecture is called wireless sensor network. In many applications, this simple system architecture is not sufficient. Furthermore, the user pays a fixed amount of money for the complete system, independent of the usage of e.g. server and services. For more complex algorithms and a location-independent application of the system, other system architectures have to be used. Computationally intensive algorithms for step segmentation, which e.g. are combined with a comparison of steps and statistics from a database, should run directly on a remote server. Furthermore, it is proposed that the cost of the system should be dependent on the usage of server and services.
The goal of the project is to develop a generic platform for future Wearable Computing trends in sports and fitness, which has the following four features: (i) permanent availability by integrated sensors in apparel using e-textiles (sensor integration), (ii) intelligent and flexibly adaptable sensor data processing for a detailed analysis of complex movement patterns, which provides more than just a classification (signal analysis), (iii) development of innovative user interfaces (Human Computer Interaction), and (iv) extension of the system architecture of wireless sensor networks with a combination of cloud computing methods (Cyber Physical Systems).
