Home Sensor Date Fusion to Support Aging in Place

Period of Performance: 09/01/2007 - 08/31/2009


Phase 2 SBIR

Recipient Firm

Cleverset, Inc.
673 NW Jackson Ave.
Corvallis, OR 97330
Principal Investigator


DESCRIPTION (provided by applicant): The aging of the U.S. population presents many challenges. The existing paradigm of care will not allocate resources efficiently as the size of the population requiring home care assistance grows. Our objective is to enhance elder independence by providing both better and more timely predictive health-status assessments and direct, real-time, recommendations and warnings. These together will lower the risk of elders remaining at home or in low intensity care settings. The objective of the Phase II research and development effort is to develop technology that can detect and track activities in the home environment and to demonstrate its usefulness in allowing elders to remain in their homes longer than is now possible. CleverSet will develop and deploy a prototype CleverSet Activity Tracker, CAT, that processes data from a robust set of simple sensors to (1) track the activities of daily living over time (2) modify these tracked activities to include uncertainty about the environment and risk to produce notifications of Events Requiring Intervention (ERIs); and (3) demonstrate the results of the models. The technological innovation of the proposed work is the application of dynamic relational Bayesian networks (DRBNs) to activities in the home environment. CleverSet's DRBN algorithms collectively referred to as CleverSet Modeler, exploit the data model and meta-data from the schema to guide and frame relational queries about behavior and events. In the proposed work, DRBNs will be used to represent complex, dynamic, multi-scale processes involving multiple actors, as probability distributions over the elements, queries, and relationships in the DRBN model. Activities of daily living (ADLs) will be identified using DRBN machine learning algorithms from sensor data and tracked through time. Short-term rhythms of daily life as well as longer-term transitions will be tracked. Risk modifiers relevant to elders will be integrated into the model and used to adapt the sensor data input. Sensor studies will also be performed to determine the relative contribution of sensors to the DRBN ADL models. A software prototype integrating the elements of the Phase II effort will be developed.