AI Could Change Everything in the Delivery of Homecare

When Dor Skuler wakes up in the morning, he is greeted by a petite companion who can track how he feels. The companion’s main task? To ensure he’s not isolated and to help keep him healthy and independent.

But this companion isn’t a real person—it’s ElliQ, an outgoing digital assistant designed to engage with older people. And Skuler isn’t a typical client, he’s ElliQ’s creator. He says technology built on artificial intelligence (AI) like ElliQ could be the solution to the impending demographic crisis in care.

“There’s just not enough young people willing and able—not to come in once a month for a home visit, but to be there all day long,” Skuler, who is CEO of Intuition Robotics, told HomeCare by phone from Las Vegas, where he was at CES 2024 (formerly the Consumer Electronics Show) to introduce ElliQ 3.0. “That’s what I think AI is really well designed to do, assuming you understand the customer extremely well and you’re able to build an AI that is empathetic and can actually build a relationship.”

Intuition Robotics says ElliQ is the first social robot to use generative AI, which makes for stronger conversational abilities, independent decision-making and a deep well of memory, allowing the device to remember everything the user tells it—from grandkids’ names to their favorite meal—and use it in future interactions.

And it’s not just a prototype for electronics shows. Intuition Robotics says several state and local aging programs have given their clientele access to the devices, including the New York State Office for the Aging and the Area Agency on Aging of Broward County (Florida), as have Medicaid Managed Care Organizations and payers such as Inclusa, a Humana company.

“Today’s older adults are harnessing the power of AI to transform their lives and embrace new technologies that bring companionship, knowledge and connection into their daily routines,” Rick Robinson, vice president and general manager of the AgeTech Collaborative from AARP, said in a news release. “As we continue to combat the epidemic of loneliness among seniors, solutions like ElliQ play a vital role, offering not just innovation, but also a lifeline to a brighter, more connected future where our aging population can enjoy the full spectrum of opportunities that the digital age has to offer.”

A New Niche

ElliQ is called a robot, but don’t picture Rosie, the housekeeper from the Jetsons. Rather, it’s a tabletop device with a built-in screen in its latest iteration—a bit larger and friendlier looking than a typical smart assistant such as Amazon Alexa or Google Home, but still closer to a lamp or computer than a human being.

ElliQ—Skuler calls the device “she”—can tell jokes, play cognitive and other games, schedule rides, give reminders about medications and convey health information to family or care teams with the user’s consent. But what truly makes it different from other digital assistants, the designers say, is its proactive nature. Rather than just sleep until called on, the ElliQ will address users unasked.

“It’s hard to build a relationship with a machine that’s waiting for you to prompt it with a command,” Skuler said. “Like, how close can you get to your Alexa when the interactions are, ‘Alexa set a timer, Alexa what’s the weather like?’”

Instead, he explains, ElliQ will greet a user by name, comment on the day and ask how they slept the night before. “‘I remember yesterday you didn’t sleep well, and we did some breathing exercises before bed—was it helpful?’ And stuff like that,” Skuler explained. “The more you share with her, the better she gets to know you and the more personalized those interactions are.”

ElliQ may be garnering big buzz with recent articles in the Washington Post, Forbes and elsewhere, but it isn’t the only device on the market positioned to help people age in place. Labrador Systems has the cart-like Retriever, which can move about a home fetching objects like food and laundry for those with mobility issues. Amazon’s Astro has autonomous navigation and can be used for home monitoring and security. It’s conceivable that robots plus remote monitoring could add up to a new kind of care for the homebound.

“The concept of an ‘AI caregiver’ may seem futuristic,” aging-in-place technology expert Lori Orlov wrote in her November 2023 report on “AI and the Future of Care Work.” “But given staffing shortages today, it is feasible and works. AI analysis is derived from cameras, sensors or voice patterns on the front lines of care, alerting staff to issues based on patient health and status and history.”

Orlov said applications include:

• Using audio and visual telehealth monitoring as virtual “sitters” at the bedsides of high-risk patients
• Supplementing in-person nursing with remote patient monitoring, especially in off hours and for things like wound care
• HIPAA-compliant chatbots to divert queries from receptionists or call centers
• Analyzing a patient’s care history and, when asked a question, formulating likely responses without a clinician having to read up on the entire social history in depth
• Sensors that go beyond basic remote patient monitoring and allow for early detection of patterns that could be cause for alarm, such as temperature, light, sound or absence of movement

“It was just one year ago that the launch of ChatGPT kickstarted the revitalization of interest and excitement about AI,” Orlov wrote. “As the initial hype and fear recede, practical possibilities become viable and accepted.”

Office Tech

Products like ElliQ are dramatic examples of how AI and other smart technology could impact in-home care in the very near future. But in reality, homecare companies are more likely to encounter AI behind the scenes as part of their back-end operations and software.

One of AI’s great strengths may be to offload or streamline routine tasks for critical workers, said Joel Dolisy, chief technology officer at WellSky. For example, he said, as many as 70% of the company’s client clinicians use an hour or more after work to catch up on charting—time that everyone would prever to avoid.

“These types of advancements can really help bring some relief to this constant run on this hamster wheel of ‘I’ve got to spend a lot of time documenting after I’ve seen the patient,’” Dolisy said. “How can we start minimizing the amount of time that is needed to document things, and can we build systems that are better at understanding what is happening during the encounter?”

The key, he said, is an approach that still relies on human oversight and judgment. The AI might work from a recording and provide a draft chart for a nurse to edit. Or it could offer someone coding two likely options to choose from rather than hundreds.

AI could also vault electronic health records past what Dolisy called “glorified filing systems” to become a sophisticated tool to be used across the continuum.

“I think that, finally, we are getting to a place where they are going to be helping the clinicians really get their job done faster—and honestly, getting better outcomes for patients because those clinicians are able to spend more quality time with those patients,” he said.

Homecare organizations may also use AI in the near future to:

• Train clinicians, sales representatives and other highly paid staff, including helping them see how to do their jobs better through customer simulations
• Train new caregivers online and help them add skills to encourage them to remain in their positions
• Work with non-fluent English-speaking staff or patients more effectively
• Triage care by determining which patients are at highest risk
• Analyze unstructured data such as medication orders or visit notes to determine patterns and even predict future care needs

Overcoming Anxieties

Not everyone is on board with AI—but the experts say some anxiety about new technology, especially one this groundbreaking, is understandable.

The first step in overcoming it is ensuring that ethical questions about AI are answered and that organizations are being transparent in their use. That’s important for patient interactions like with ElliQ—which is designed to frequently remind its users it is not human—and also to watch out for potential bias or errors (sometimes called hallucinations for AI when not grounded firmly in fact) in health data.

Dolisy said that organizations should look for early adopters who are willing to experiment with AI now, but that ultimately, change will come very quickly once people realize how well the technology works. He compared it to the financial industry initially being leery of computerized spreadsheets in the 1980s.

“How many accountants are not using spreadsheets nowadays? Probably zero,” he said. “I think that’s the journey we’re going to be on. We are paving the road at the same time that we’re trying to understand what the art of the possible is. And it’s very encouraging.”