$790,906
Smart Home-Based Health Platform for Functional Monitoring and Intervention
Washington State University
Focus: To create a "smart environment" that enables older adults or individuals experiencing cognitive or physical limitations to function independently and remain in their own home setting.
By 2040, nearly one quarter of the U.S. population will be 65 years of age or older and many will require some type of monitored living arrangement. A "smart environment" is one that is equipped with low-cost sensors that yield alerts and reminders for the elderly and disabled to ensure they are adequately caring for themselves. In a specifically-built "smart home" environment, the research team will use specialized sensors to monitor whether activities of daily living (ADLs) are being adequately performed and whether ADL monitoring provides a basis for effective interventions using reminders and automated assistance. In 2005 Washington State had several hundred thousand individuals being cared for in nursing homes at a cost of about $94 million per day, not including assisted care or veterans facilities. If one percent of the state's population is kept at home for two additional years, the result will save Washingtonians approximately $9.4 million a day and $3.4 billion a year.
Organizations collaborating in this research are: Group Health Cooperative, Rosetta Inpharmatics LLC, and Children's Hospital & Regional Medical Center.
Grant Update
“We have created the world's largest repository of smart home data as part of this project. All of the data is annotated with activity levels. We have tested our activity recognition algorithms on all of these datasets. In addition, we have created a supervised learning algorithm that learns general models of activities based on multiple datasets and uses the setting-generalized models to recognize activities in new environments with new residents. The recognition rate for the setting-generalized approach is almost the same (an average of 3% reduction in accuracy, significance p<.6) as when training data is available for each specific environment setting. In terms of the audio cues, we have demonstrated that these cues make a marked difference in the ability of older adults with dementia to independently complete ADL activities.”
See also:
