Turning Domestic Appliances Into a Sensor Network for Monitoring of Activities of Daily Living

摘要

Smart sensors provide a support for turning a typical household Into a technology fitted environment for elderly or disabled who need tracking or assistance with activities of daily life. This paper presents a design of in-house infrastructure of sensorized appliances making a measurement context for long-term studies of behavior and early detection of health setback. Each appliance-specific sensor set reports a multimodal data on the position and usage of the device as well as operating people. Besides seamless security watching, these data are integrated in personalized behavioral records (actigrams) specifying activities of daily living. This record is then automatically explored to identify patterns of circadian sequences to assess a lifestyle and provide a reference for detection of abnormalities. The purpose of newly proposed monitoring system is twofold: increasing the security of independently living elderly and identifying their habits and interpreting inhabitual actions as a possible health setback. The system uses remotely configurable sensors detecting human activity by collecting simple information in a pervasive, imperceptible way.