Skip to Main Content
Within the project "eHome" a prototype of an assistive home system was developed, aiming to prolong the independent life of elderly people at home. Besides communication, e-access and safety relevant features, a core part of this system is an automatic fall detection, which utilizes floor-mounted accelerometers to gather body-sound signals that typically occur during a human fall. This approach targets to avoid acceptance, usability and reliability issues of available body-mounted fall detectors. The system was developed with focus on practical applicability, reliability and exploitability. The prototype was evaluated successfully in laboratory and during 507 days in real-life at homes of persons from the target group. During the laboratory trials a sensitivity of 87% and a specificity of 97.7% could be achieved for a defined fall scenario and across four tested floors. Further research is suggested to investigate floor dependencies of the fall detection performance.