WOAR 2014, Workshop on Sensor-based Activity Recognition
11. März 2014, Rostock Warnemünde; Proceedings
Sensors worn at the body allow an unobtrusive recording of physical activities, of tranquillity, sleep and stress, and thus support the trend of quantified self. Through MEMS components (Micro-Electro-Mechanical Systems) it is possible to employ a great number of electronic devices to accompany everyday activities. They are of interest not only for private users but also in the context of industrial applications for a continuous monitoring of life and work situations. Human activity recognition is the sensor-based, mostly unobtrusive and continuous recording of physical activities, its analysis and user-related application. It is an interdisciplinary field of research with several technical challenges, and it also encompasses topics from medicine, psychology or industrial science. As the sensors only give a simplified image of reality it is necessary to analyse the data and to place them in the appropriate context. Therefore, innovative solutions are needed for sensor technology, preliminary data processing and machine learning as well as for new human-machine interfaces and assistance technologies in the respective fields of application. The unobtrusive monitoring of people with the help of very small electronic systems attached to the body makes it possible to give complex support in the fields of medicine, profession and leisure. Miniaturisation, new algorithms and concepts open up new fields of application for the recognition of activity at or in the body and thus to be able to assist people at any time and any place. This requires a new understanding of personal assistance and of human-machine interfaces. The first workshop on sensor-based activity recognition in Rostock-Warnemünde, WOAR 2014, brought together scientists, interested parties and users. It provided an opportunity to exchange experiences and present best practices as well as technical and scientific results. The participants dealt with different technologies for the recognition of physical activity with the help of inertial sensor systems (acceleration sensors, gyroscope etc.) and also their practical application. The present publication contains the contributions to the workshop. They reflect the current state of science and technology concerning specific technologies and applications in the fields of (occupational) health, lifestyle change and quantified self as well as production, maintenance and service.
Fraunhofer-Institut für Graphische Datenverarbeitung -IGD-, Institutsteil Rostock