OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign up, Login or Learn more

Decrease worry and improve care management with information from data analytics

Caregivers get peace of mind when sensor analytics provide oversight and understanding of a care recipient at home alone.

Photo of Laura
0 2

Written by

Who is your idea designed for and how does it better support family caregivers as they care for a loved one with dementia?

EmPowerYu is a Care Management tool ideal for optimizing care delivery to persons who cannot accurately report what is happening at home from day to day. EmPowerYu reassures caregivers when home patterns are normal, alerts them to problems, and measures the effect of treatments better than surveys, phone calls, or occasional visits. Knowing what is happening in the home, and having a system that will call out for help, reduces the stress and worry of a working or long-distance caregiver.

One of the greatest concerns of caregivers is not knowing what is happening in the home when the care recipient is alone, particularly if he or she is at risk for falls or medical complications.  Six in ten dementia caregivers were working an average of 35 hours per week, 57% of them full time (Merrill Lynch/Age Wave The Journey of Caregiving report 2017) separating them from the care recipient for many hours per day. Also, misinformation from the care recipient who does not remember accurately when answering a question can mislead caregivers.  Along with reducing worry, accurate knowledge of events in the home can make caregiving easier by providing independent information about wellness, indicate changes in condition, and show effectiveness of treatments.  For example, at-home caregivers can better understand and manage sundowning by seeing the night time activity of the care recipient when the caregiver is asleep. The caregiver can test which interventions work and which don’t, thereby improving both people’s sleep. (See the WebMD article Sundowning,: Cause, Symptoms, and Treatments at; the article recommends motion detectors and door sensors that report night-time activity.) 

Monitoring can also record what is happening during the day, particularly when a family caregiver is away at work.  One middle aged couple used EmPowerYu to see the mother’s activity in the home during the day when they were at work – the mother liked a late morning bath and the couple checked at lunchtime to see that there was kitchen activity, indicating that the mother had safely bathed and dressed. A family using EmPowerYu can see that a paid caregiver arrives, prepares a meal, and leaves at the contracted time. The care recipient can’t always remember that the paid caregiver had been to the home, so this ensures that the family and the caregiver company have independent confirmation of services provided.

What is the best way to monitor a person with dementia?

  • A video gives good information, but cameras are invasive and disturb people’s privacy.
  • Wearables like smart watches and fitness bands require active participation to get the data, including charging the devices and putting them on consistently. 
  • The market research on personal emergency buttons and FitBit activity monitors shows that only about 10-15% of people continue to wear these devices after a few months, because using them requires behavioral change, and the data gets monotonous when things are ‘normal’.  So longer term monitoring should not rely on cameras, wearables, or behavioral changes.
  • Many wearable health and fitness devices require data upload with a smart phone, which many older adults don’t have, particularly low-income seniors who may greatly benefit from improved oversight. 

EmPowerYu uses familiar home security and automation devices that are installed in strategic locations on the walls of the home to passively detect patterned activity related to eating, sleeping, moving around, and use of devices like TVs and coffeepots. These human-initiated events include motion in a room, turning on the TV, opening the refrigerator, using the microwave, etc.  The home resident does not have to do anything differently, and EmPowerYu’s users report that they don’t even notice the few visible sensors after a week or two.

EmPowerYu’s analytics compare data from the current day to each person’s own ‘normal’ patterns. Long distance caregivers get reassurance that conditions are normal, and at-home caregivers can better understand what is happening at any time. Families participating in remote monitoring report a reduced emotional and physical burden of caregiving (Alberta Health Services: Clinical Outcomes (2016; from Care Innovations)).  Mathematically describing the normal boundaries of personal behavior enables us to detect immediate problems, like not getting out of bed at the usual time (possible early marker for depression or sickness), unexpected quiet in the home that could indicate a fall, lack of meal activity in the kitchen, long periods of activity at night, or too many or too few trips to the toilet. We could also do trend analysis to look for an impending problem, like slowing gait or sudden weight gain. Understanding each person’s patterns with behavioral modeling and machine learning gives family and professional caregivers an objective care management tool to understand each care recipient’s complex and unique situation.

In addition to ongoing wellness and safety monitoring, caregivers are particularly concerned about an older adult who returns home after treatment at the hospital or a nursing facility.  A retrofit and easily removed sensor system can also provide short-term oversight at this time of particular vulnerability, documenting important parameters and showing trends to remote caregivers.

A quantified measurement of the Activities of Daily Living can show effectiveness of treatments.  For example,

  • Does restricting coffee to morning affect night time activity?
  • Confusion is often detected from repeated opening and closing of the refrigerator door without follow-on meal preparation. Does a new medicine or routine reduce this repetitive activity? That may be hard for a caregiver to judge, but EmPowerYu can report the number of times the refrigerator door is opened and closed.
  • Restlessness can be detected by prolonged room activity; does a daily walk reduce restlessness and increase sleep time?  

For healthcare providers, individual anonymized datasets can be combined into a database for outcomes-based research into the broader effectiveness of interventions in specific groups, improving quality of care for all. Many older adults, including persons with dementia, have more than one chronic condition, so a care coordination tool that measures both high level and specific behaviors is a holistic approach to understanding and improving the overall wellness of the care recipient.

Specifically, we:

Collect data – a minimum number of strategically placed passive home security sensors get information on human-initiated activities and events related to eating, sleeping, daily activity, and use of devices like a TV or coffee pot or lamp.  EmPowerYu is an open platform, we can incorporate any data we receive, from home sensors to medical devices to environmental readings like air quality.  We get the best sensor on the market for the data we need, rather than being restricted to proprietary sensors.

Analyze the data – our first step is behavioral state modeling, then we overlay machine learning and an inference engine that categorizes events like morning rise time, gait, meals, etc. to create a personalized Activity Profile™.

Measure / compare

  • daily data is compared to the historical record for that person to understand ‘normal’ conditions, and
  • data can be cross-correlated to discover or monitor relationships.

For example, walking speed can be tracked according to time of day, compared to previous days, or to an average over the last month or the prior year.  Walking speed can also be compared to the prior night’s sleep duration, to amount of night-time activity, to the weather (atmospheric pressure or air quality), to a medical device reading, or any other metric for which we can get the data.  

In addition to direct detection of events like getting up in the morning, our inference engine could be set to detect symptom clusters that point toward high level wellness indicators like fatigue or nausea. These high level indicators often precede more serious problems.  For example, the day(s) before having the flu, we feel slow, tired, achy, not interested in food.  When a care recipient doesn’t get out of bed within their normal range one morning, EmPowerYu can look for symptoms the previous day, like an early bedtime, slow walking speed, or unusually low meal preparation, to associate with a late rise and reinforce concern for the person’s wellbeing.  This could compare to a late rise preceded by having the TV on unusually late the night before, and a late bedtime, which would decrease the urgency of a response to a late rise.

Provide a dashboard for information and alerts for intervention – our user-friendly interface shows current status on the top page, with increasing information available on lower pages. At lunchtime a busy family caregiver can see at a glance whether everything is normal at their care recipient’s house because a big green checkmark means the care recipient has gotten up within the normal time, eaten a morning meal as usual, is moving around, and (if applicable) has watched their usual morning TV show.   The Settings page allows a caregiver to select notifications for normal events, like when a care recipient gets up in the morning.  There are also Alert Settings for out-of-normal behaviors and out-of-range readings. Out-of-normal behaviors might include failing to get up in the morning, not having a morning meal, or lack of activity for 6 hours when the person is at home (possibly a fall?).  Out of range readings might include too high or too low home temperatures or blood pressure measurements.

                (Optional) Validate alert before sending it to designated caregiver – if requested, we can provide a low cost high value auxiliary service that receives EmPowerYu alerts, and calls the care recipient to confirm the alert before contacting the designated caregiver.  This ensures that caregivers get timely and relevant alerts, and are not bothered by unusual but unimportant changes in home patterns.

Report – (in process) for professional in-home care providers, actions and analyses from the home can be tabulated in a report for daily or monthly tracking of individuals, then collated for analysis of group outcomes and statistics. EmPowerYu is building a Big Database of cross-correlated daily activities, environmental measurements, and home medical device readings that will be a powerful outcomes research tool.

                Also, families often employ home care agencies for tasks like housework, bathing, or meal preparation.  A monitoring system can detect when a paid caregiver arrives, whether there is appropriate activity in the home, and whether the paid caregiver stays for the expected time.  A person with dementia may not be able to confirm these activities accurately, so an independent reporting system protects both the family hiring the paid caregiver, and the staff providing the service, by documenting the visit.

The benefits of EmPowerYu’s service include:

  • Real time assurance that the care recipient is OK, which decreases stress of family caregivers; knowing that the home pattern is normal is a mini-respite from worry
  • Provides standardized information that can be shared with others in a group of caregivers (potential tool to reduce caregiver isolation?)
  • No cameras, wearables, behavioral changes, or technology that the care recipient has to learn to use
  • Unobtrusive passive monitoring that maintains dignity because nothing touches the body, creates a visual image, or invades privacy
  • Provides both short-term and long-term oversight as needed
  • An open platform that has a standard sensor set for safety monitoring but expands for customized applications
  • Holistic monitoring to improve overall health and safety, enabling Aging in Place which provides psychological comfort to the care recipient and preserves financial assets of the family
  • Low cost, high value service that can be shared by families, friends, and healthcare providers
  • Healthcare providers can get reimbursement through Chronic Care Monitoring codes
  • Low cost commodity hardware so service is affordable
  • Keeping an older adult in his or her own home preserves family assets compared to moving to assisted living ($80-90,000 per year in  nursing home compared to less than $45,000 for in-home services for persons who do not need round-the-clock care); reducing financial burden also decreases family stress

“EmPowerYu has the flexibility of setting up unobtrusive devices which monitor my Dad’s personal activities of daily living…I can be watchful without being invasive.”  Linnea, long distance caregiver of her 94 year old father.

“I find it very helpful, thank you!” Jenny, daughter of separated parents, each of whom have an EmPowerYu system in their homes.  Before EmPowerYu, each parent left daily phone messages. Jenny panicked if they forgot to call, and then weren’t home when she called later. With EmPowerYu, she can see when there is activity in the homes.  If they don’t answer the phone, she can see if they are out of the house and not worry that they have fallen and are on the floor (which happened before EmPowerYu was installed).

What early, lightweight experiment might you try out in your own community to find out if the idea will meet your expectations?

We have systems in homes already. One test home is a full time caregiver of his aunt with advanced dementia. He likes to see what is happening at night when he is asleep, so that he can interpret his aunt’s daytime activities with the information about whether she slept well the night before. He also wants to use EmPowerYu’s front door sensor as an alarm so that if she leaves the house at night we can call his cell phone and wake him up to go get her. We will experiment with this soon.

What skills, input, or guidance from the OpenIDEO community would be most helpful in building out or refining your idea?

We can hypothesize many possible advantages to specific data collection, as given above. However, collaboration with the OpenIDEO community to brainstorm ideas for what data could help caregivers, refine the ideas, then prioritize the ideas with real caregivers, would be terrific. We want to work with caregivers to validate the usefulness of these ideas, and connect an alert with next steps like a televideo conference with a behavioral expert, music therapy, or talk with an avatar.

How long has your idea existed?

  • Over 1 year

This idea emerged from

  • An Individual

Tell us about your work experience:

The EmPowerYu team members have all been caregivers for their parents, have prior startup experience, and design for data security. Laura has worked in healthcare, Rich has won awards for human-centered user interface design, and Jeff ‘s expertise is signal analytics and safety-sensitive design.


Join the conversation: