CareWheels Participatory Design Research

I have been able to find a part of myself
that I was told was gone, because of the stroke.
 
Laura B., CareWheels Participant, Mother, Grandmother, Stroke Survivor
 

Social Technologies for Interdependent Care

CareWheels explores current opportunities to apply social technologies that positively impact the Social Determinants of Health at the intersection of three convergent megatrends: the Age Wave, Information Revolution and Financial Instability. This social health perspective is important now because it de-emphasizes the over-medicalization of natural aging processes and focuses instead on the low-hanging fruit that are within our reach to collectively address the consequences of the emergent Age Wave demographic megatrend. Emerging gerontechnologies like Telecare may be harnessed to empower Boomers, elders and caregivers with a hybrid peer-care model, based on reciprocal interdependence, to drive innovation adoption. Evidence-based Telecare services have the potential to improve our quality-of-life and reduce health care costs by reducing preventable emergency room visits and hospital admissions, using a variety of means including technology-empowered self-care, early disease detection and prevention, and improved chronic disease and medication management. Social technologies and mobile devices empower us to provide a broad range of medical, health and social services that support our desire to age in place, just in time to meet the urgent needs of our aging society.

Download: Social Technologies for Interdependent Care.pdf

 

Participatory Design

CareWheels was founded in 2001, with grant support from the Intel Research Council, to develop and evaluate a set of smart home technologies at the Pine Point Apartments in Portland, Oregon, a barrier-free building for independently living persons with severe disabilities. Our grant application proposed using participatory design methods to engage pre-senescent people with severe disabilities as proxies for frail elders, who have similar, but generally more stable cognitive, motor and sensory challenges. They have helped to guide our research, co-development and evaluation of in-home sensor systems. 

Goodman CA, Jimison HB, Pavel M; Participatory Design for Home Care Technology, Proceedings of the 2nd Joint Conference of the IEEE Engineering in Medicine and Biology Society and the Biomedical Engineering Society; (2002)
 
The Pine Point Living Lab demonstrated a novel research method that combined Participatory Design with Remote Usability Testing in real people's homes. Data collected by the Living Lab was shared freely with academic and industry colleagues and the wireless sensor methods that CareWheels developed were subsequently deployed by the Oregon Health & Science University's ORCATECH Living Laboratory and the Intelligent Systems for Assessment of Aging Changes Study.
 
ParticipatoryDesignMethod
 
 

Internet-enabled Assistive Technologies for Independent Living

Sponsoring organization: Intel Research Council                                                            

The goal of this 3-year annually renewable, merit-based grant award was to develop and evaluate a set of smart home technologies including, a wireless sensor network, videophone and local social network to demonstrate function, integration and utility with a group of pre-senescent persons with severe disabilities serving as proxies for frail elders. A novel research method was used that combined Participatory Design with Remote Usability Testing. Collected data sets were labeled, anonymized and shared freely with research colleagues and our wireless sensor implementation was subsequently deployed by Oregon Health & Science University in the ORCATECH Living Laboratory and the Intelligent Systems for Assessment of Aging Changes Study. A review of the Pine Point Research is available:

CareWheels Research Review: Internet-Enabled Assistive Technologies.

 

Pine Point Project

CareWheels implemented a residential Living Laboratory at the Pine Point Apartments to develop and evaluate a set of smart home technologies including: a wireless sensor network, videophone and local social health network to demonstrate function, integration and utility with a group of pre-senescent people with disabilities serving as proxies for frail elders. The Intel Research Council produced this CareWheels TeleCare Project video, in which Eric Dishman, General Manager of Intel's Health Research & Innovation Group, explains:

The magic of CareWheels is taking these two disenfranchised communities: people who are of working age with physical disabilities, and using technology to help them support frail elders who are trying to live at home in the community. With a program like CareWheels, we've seen successful use of wireless sensor networks, the successful creation of interfaces that both frail elders and people with disabilities can use.

 

SmartHome Inference Research

  

PinePointApt1 CWQC Sensors-web

  

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:

Situation
Data
Needs
Preferences

A. Bathroom fall with
 fear of being ignored

Missing meds & motion data after fall

Automatic anomaly detection and alerts

No need to wear her call-for-help button

B. Diuretic effect of a
    new medication

Changed bathroom trip frequency data

Detection of changes induced by meds

Objective data to show her physician

C. Forgetting to take
     her medications

Missing data for meds cabinet access

Conditional prompts to support her lasting competencies

Reminding only when she actually forgets to take her meds

 

Home Sensor Data Fusion to Support Aging in Place

Sponsoring organization: National Institutes of Health, Phase-I & II SBIR                                                             

The goal was to build a prototype activity tracker that would detect and track a subset of activities of daily living, create alerts about meals and medications, and to infer changes in the health status of an individual over time. A living laboratory was implemented in the apartments of participants to collect sensor data. Dynamic Relational Bayesian Network methods for fusing sensor data were used to infer behaviors and calculate the risk of falls over time to provide information and support for elders living independently.

Goodman CA, Jorgensen J, Rossignol AM, Paton L, Jensen S; Technical Report: Home Sensor Data Fusion to Support Aging in Place, funded by National Institutes of Health, Behavioral Science Research Division, Small Business Innovation Research Phase 2 grant number 5R44AG02468-03; (2010)
 
Excerpt: “Co-author, Claude Goodman, is expert in discerning activities from sensor streams. In particular, his expertise is based in his knowledge of the physical capabilities of the sensors, how they are placed, their reliability, the physical capabilities of the resident in the sensored residence, and ways to quickly process the data to obtain sensible extracts to support his inferences. One goal of our project was to automate this expertise.”
 
Domain experts are familiar with the households, instrumentation and methods by which sensor data was collected. We apply this 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 wellness monitoring in the home.