CapBed - preventive assistance system for the bed area based on capacitive sensing

Over the past decades, human activity recognition systems have become a major input modality for building automation. However, those systems also found recent applications in emergency detection, such as recognizing patient activities that may lead to life-threatening situations like falls or heart attacks. The aim of this thesis is to develop a sensor that recognizes whether a person wants to get out of bed. This is to prevent falls by illuminating the path or calling a nurse in time. In addition, such a system can also provide insights into the behavior of the user in the long term. Therefore, a concept of preventive assistance system for the bed area based on capacitive sensing is developed within the scope of this work. To this end, a comparison to other sensor technologies will be established, followed by a detailed overview of the technical background of capacitive proximity sensing. An innovative concept of a device that offers decent performance at an affordable price is proposed. Based on this concept a prototype system was developed and evaluated to investigate its sensing performance and identify possible limitations. As a future outlook, this thesis summarizes the occurred problems and suggests possible modifications that might improve the overall performance of the system.