A smarter home could ease the strain of dementia care

Living at home can bring daily uncertainty for family and caregivers of people with dementia. About 80% of people living with dementia in the United States live in their own homes, and 89% rely on informal caregivers such as spouses, adult children, other family members and friends. Safety risks often push families toward stricter supervision or major care decisions that can limit a person’s independence. 

In Texas, about 400,000 people are living with dementia, according to Dr. Marcia Ory, regents and distinguished professor in Texas A&M University’s School of Public Health. 

“Most people want to live at home as long as they can, and those with dementia are no different,” Ory said.

Texas A&M University researchers are developing a smart home system to help people living with dementia stay safer at home and alert caregivers to problems such as falls, wandering and kitchen hazards. The system could gather information from home sensors, smartwatches and other connected devices, then bring data together in one place — the Smart Home Care Digital Twin app — so caregivers can monitor safety and respond when something seems wrong.

“Interestingly, about 20% of people with dementia live at home by themselves, making smart home apps a critical technology for supporting independent living,” said Ory, a co-investigator on the project

The project, funded by the Texas A&M Health Dementia and Alzheimer’s Research Initiative (DARI), brings together faculty from architecture, engineering and public health to address a growing challenge in dementia care.

Ory serves as one of DARI’s three co-leads and said the project stood out from over a hundred submissions because of its practical impact. “It’s with real people, and it’s making a difference now, not 20 years from now,” Ory said.

Smart home technology for caregivers

Many homes already have cameras, smart appliances or wearables that can collect information useful for caregivers. Researchers say the problem is that caregivers may not be familiar or comfortable with these technologies, and if they try, they have to sort through multiple apps and alerts when they may already be overwhelmed.

Dr. Xuemei Zhu, endowed professor in the College of Architecture and co-principal investigator, said the team wants to close the gap between what technology can do and what families actually need and can adopt easily in everyday dementia care.

Zhu said a garage door repair taught her an unexpected lesson. She initially resisted downloading the app for her new garage door because it felt like just one more app to manage. However, unlike some underused technologies, this app provided exactly what she had been needing — the ability to check on her phone if the door was closed after leaving home or when getting ready for the night.

The experience shaped how she thinks about this project. Zhu said that the right fit and ease of use are key to the success of smart technology. Researchers can build sophisticated technology, Zhu said, but it will not help if families do not find it practical and easy to use.

A digital twin built for home safety

The project will develop and test a phone app based on a “digital twin” of the home. 

The app can combine data from home sensors, including radar sensors for presence and falls, door and window sensors, vibration sensors, indoor environmental quality sensors, smoke alarms, smart meters and appliance monitors. It can also integrate data from smartwatches and video cameras. 

Working together, the system will monitor a person’s location, behavior and health, detect safety risks such as falls or a stove left on and alert caregivers when needed. 

If a sensor detects a fall, the system could alert the caregiver and combine that information with data from other sensors or cameras to give caregivers more context about what happened. If the entrance door opens at an unusual time — like at 3 a.m. — or a stove stays on longer than usual, the system could flag it as a potential hazard and alert a caregiver.

Researchers also plan to compare smartwatches with millimeter-wave radar sensors to determine which works best for fall detection in home-based dementia care.

Co-principal investigator Dr. Zheng O’Neill, endowed professor in the College of Engineering, said the team will use Home Assistant, a widely used open-source smart-home platform, as the core system in this pilot study because it allows flexible, local control of many different devices. This choice keeps the system affordable while supporting privacy‑preserving automations and customizations. 

The team also plans to ensure the plug-and-play capacity of this smart home platform so that families can customize it for different caregiving needs.
“The innovation feature of this project is not just adding sensors,” said co-investigator Dr. Junhyoung (Paul) Kim, associate professor in the School of Public Health. “It is integrating disconnected smart technology into a unified digital twin that mirrors real-time home environments for dementia care.”

Shaping the app with caregiver feedback

The 15-month effort began in January and is still in its early stages. Zhu said the team is working on the initial field tests of sensors, devices and the Home Assistant platform in the Texas A&M Smart and Connected Homes Testbed at the RELLIS campus. The testbed site includes two 1,200-square-foot lab homes with removable walls that let the team simulate different home conditions by changing layouts and moving sensors.

The team is also interviewing caregivers and people living with mild cognitive impairment or dementia to identify the most desirable features from the users’ perspective. 

Researchers said these conversations will be central to the system’s design, helping them understand what technologies caregivers already rely on and what features would make the system easier to adopt. 

Kim said the larger goal is to use technology in ways that reduce the burden on both people living with dementia and their caregivers. 

Student participants will simulate daily activities while wearing aging simulation suits that mimic sensory and physical limitations associated with aging. Researchers will also recruit judo and wrestling athletes, who are experienced in controlled falls, to simulate fall scenarios safely. 

The setup allows researchers to test high-risk situations, such as falls when getting out of bed or using the bathroom, without putting older adults in danger. 

Researchers will evaluate how well the system performs and how usable caregivers find the final app. If the pilot is successful, the team plans to seek outside funding to test the system in the homes of people living with dementia. 

By Ana Renfroe, Texas A&M University College of Architecture