Home Sensor Data Fusion
Home Sensor Data Fusion to Support Aging in Place
a Phase-1 SBIR sponsored by the National Institutes of Health
Jane Jorgensen and Bruce D’Ambrosio, CleverSet, Inc. Claude Goodman, CareWheels Corporation
Abstract
CleverSet, in conjunction with CareWheels, has developed an innovative approach to monitoring and assessing human behavior from sensor data. The objective of this work is to evaluate the feasibility of using Dynamic Relational Bayesian Networks (DRBNs) for functional assessment (in terms of activities of daily living) and anomaly detection (e.g., falls). We have progressed from relatively coarse sensor-level models to more refined dynamic hidden relational models that relate sensor data to a home care ontology.
Research Objectives
To fuse wireless sensor data to monitor living behaviors (such as bathing, dressing, eating, etc.) of occupants in an assisted living setting. The sensor data have been collected by CareWheels at their home sensor testbed currently operating at Pine Point Apartments, an independent living facility for people with severe physical disabilities in Portland, Oregon.
Data for modeling was provided by the CareWheels Pine Point research project, “Internet-enabled Assistive Technologies for Independent Living and Aging-in-Place,” funded by the Intel Research Council. Our method engages pre-senescent persons with disabilities as proxies for frail elders in iterative participatory design to explore emergent issues of usability, privacy, safety and reliability in their networked and sensored homes.
A white paper describing the outcomes of the Phase-I research is available: Home Sensor Data Fusion To Support Aging In Place.