Matthies, Denys J.C.Denys J.C.MatthiesHaescher, MarianMarianHaescherChodan, WenckeWenckeChodanBieber, GeraldGeraldBieber2022-03-1527.10.20212021https://publica.fraunhofer.de/handle/publica/41265510.24406/publica-r-41265510.1145/3453892.3454001In this research, we present a proof-of-concept Do-It-Yourself (DIY) bed sensor mat that is capable to detect five user postures, including: Supine position, Prone position, Lateral position (left), Lateral position (right), and Sitting. Our prototype incorporates two types of sensors: Pressure Tiles and Capacitive Sensing. To gain an impression on the accuracy, we conducted a study with 11 participants and developed a machine learning model showing accuracy of 85% for an unknown user. In fact, nocturnal postures extend to a third of ones lifetime and therefore, a posture detection can be particularly interesting. Enabled application scenarios include pressure ulcer prevention, bed-exit intention detection, diabetes detection, and sleep apnea mitigation.enhardwareLead Topic: Individual HealthResearch Line: Human computer interaction (HCI)smart devicesapplications for pervasive computinguser intention analysis006conference paper