CareWheels

Research and Development

CareWheels R&D is based on two premises:

Pine Point Project

CareWheels implemented a residential SmartHome test-bed at the Pine Point Apartments to iteratively refine sensor-based embedded assessment applications and data collection methods. To help advance the research, we have provided sets of exemplary data to stimulate new trans-disciplinary participation on the problems of aging-in-place.

The value of such databases is greatly enhanced when it is accompanied by domain expertise: access to those researchers who are familiar with the households, instrumentation, data and methods by which it was collected. Domain experts apply their knowledge to the tasks of identifying meaningful behaviors and events in the sensor data stream, verifying that the discovered behavioral models represent underlying causality, and helping to transform those models into actionable programs that can provide intelligent assessment and assistance in the home. The resulting programs provide the logical and inferential drivers for innovative applications of context-awareness, embedded assessment and remote wellness monitoring. A review of the Pine Point Research is available: CareWheels Research Review: Internet-Enabled Assistive Technologies.
 

SmartHome Inference Research

There is growing recognition in the gerontechnology research community of the potential value of repurposing existing SmartHome technologies, such as wireless computer and sensor networks, to provide cognitive and social supports for households dealing with the challenges of dementia. Until such time that effective means exist for the treatment and prevention of cognitive disorders, such as Alzheimer’s disease, technological assists may offer the most compelling adjuncts to compassionate care. For example, in early to moderate stages of memory impairment, context-aware assistive devices may interpose intelligent reminders to help support task completion within the familiar and supportive environment of one’s own SmartHome – without undermining the person’s remaining competencies. Over time, as tasks shift to the caregiver, these same devices could generate embedded assessments of the remaining abilities to provide an ongoing realistic measure of what the person can do, keeping expectations realistic to reduce the stress and frustration of everyone involved.

Elders' needs and preferences may be gleaned from a real-life scenario:

A. Last month, after a bathroom fall, Lily’s caregiver had a personal emergency response system with a call-button and speakerphone installed. Lily finds it uncomfortable to wear the call-button pendant all the time and worries that it might be out of reach next time she really needs it.

B. Lily feels run-down but she has no idea why. Last week, her physician prescribed a new medication with known diuretic effect. Although Lily is not consciously aware, her nighttime trips to the bathroom have more than doubled.

Lily recently enrolled in a CareWheels TeleCare trial to evaluate a new embedded assessment system designed to:

• monitor her medication schedule adherence and nighttime activities,

• compensate for poor sleep quality by reminding her to be vigilant,

• prevent her from forgetting to take her meds, and possibly from falling,

• alert her telecaregiver in the event of an event requiring intervention.

Now when Lily arises after a fractured night’s sleep, she notices a change in the display of her CareWheels system, which makes her aware of this fact and reminds her to be more vigilant.

C. Additionally Lily receives a prompt on her PC screen reminding her that she has missed a dose of her medications. When she subsequently takes her meds a sensor in the medicine cabinet door sends a wireless signal to her PC, which automatically ends the prompt. Should Lily fail to respond to these prompts, her telecaregiver, Ann receives an email alert and then calls to check on Lily. In the case of this example where Lily’s forgetfulness is caused by her frequent night awakenings, Ann might suggest to Lily she call her physician, who may query the SmartHome assessment database regarding this particular side effect of Lily’s latest prescription, to assist him in evaluating the situation. Had Lily not answered the phone, Ann, the telecaregiver would have intervened by dispatching Lily’s designated responder to her home.

Note that this powerful integration of meds conformance monitoring, sleep quality compensation, anomaly detection and alerts is an example of the inherent synergy of data fusion and the power of inference across multiple sensors. The scenario is summarized in the following table:

^ Top