Can Technology Help Solve Our Elder Care Crisis?

While (at least so far) a robot can’t cook meals or provide hands-on care, can AI help in other ways? Chia Lin-Simmons says yes. She’s the CEO of LogicMark which creates medical alert devices. They are now using AI to enhance these devices to include precise multi-sensor detection and predictive, proactive monitoring using longitudinal data and pattern analysis to anticipate risks.

Chia-Lin explained all of this to me in a recent Risking Old Age in America podcast. Here are some excerpts from our conversation:

Risking Old Age: You told me that your experience with your mother-in-law had a lot to do with your approach to your work at LogicMark. Please explain.

Chia-Lin Simmons: She came to visit me. I was working at Google and we decided to take a quick lunch at a ramen restaurant in Oakland. Unfortunately, my mother-in-law sat down a little too quickly and her medical alert alarm went off. It immediately screams out, “Mrs. Becker, are you okay?” And of course, everything went silent in the restaurant and turned and looked at her. She was mortified and I felt so awful for her.

She’s this beautiful, elegant retired art teacher and she had this really awful, weird-looking beige lanyard around her neck that looked like a garage opener. I thought, “Gosh, it’s not very precise, and also it’s really ugly, so somebody should do something about this.”

ROA: So how are the LogicMark products different from the devices in those old “I’ve fallen, and I can’t get up” television ads?

Simmons: Our aim is to truly transform it from what I would call a reactive technology. We need to apply better optimized technology to get people help when they need it. We all know if you get help within a “golden hour,” as they call it in emergency room, your health outcomes is better.

But the reality is that, as I’ve mentioned through my mother-in-law’s story, these things from a fall detection perspective aren’t very precise. They’re running on an accelerometer, which is purely running on speed.

So a lot of people ask me, “Wouldn’t an iWatch be okay for my senior father-in-law or mother-in-law?” My response is that when you’re trying to create a device that’s going be great for a 10-year-old all the way to a 90-year-old, you’re going have to compromise somewhere. For most people, if you’re moving at a fast velocity and then you crash or you fall, that really works.

But the reality is that this is not the falling experience of most aging population folks. They slump. They experience a slow fall that the device doesn’t always capture. So our goal is to optimize the reactive technology so there’s more than one sensor that’s being used so that we can be more precise about understanding and capturing a fall versus everyday active movement.

We don’t want it to do more false alarms, because one reason why people don’t want to wear them is because it’s embarrassing if it goes off unnecessarily.

But really what differentiates us most is this idea of looking at predictive and proactive technology, which is what can we do to make these devices smarter. We utilize their capabilities to help us capture longitudinal data so we can actually try to get ahead of that fall instead of just waiting to react to that fall. That’s where medical alert products, remote patient monitoring and where the future of aging tech should be heading — to get ahead of an incident versus just react to it?

ROA: In Boston near where I live, there’s a company called Whoop that monitors people’s vitals for younger people who want to maximize their athletic ability and improve their health Is that similar to what you’re talking about?

Simmons: Products like Whoop and iWatches are used to do what I call the “quantifiable self” today. We’re monitoring our heart rate and how many steps we take and all of that. That’s really cool, but all that data sits in pods everywhere, and they don’t really interact with each other.

What we’re really interested in is trying to get ahead of incidents by gathering of those data and having AI do what it really does best, which is looking at pattern changes across a long-term sort spectrum. Whoop and iWatch are capturing that data, but they’re not necessarily looking for pattern changes that might have indications.

So if you were a 5,000-step walker and suddenly they’re slowly going to 4,000, 3,000, 2,000 steps. What’s happening? Or, for example, you usually wake up at 8:00 and go to sleep at 10:00, but slowly you’ve been shifting your pattern towards waking up at 10:00, going to sleep at 8:00, what does that mean?

We actually built into a digital twin of my mother-in-law, Mrs. Becker, which we compared to an aggregate of people with a similar background — the meds that she’s taking, age group, etc. With that, we could say, for instance, that there’s a possibility that you could fall in nine months, and so what do we need to do to get ahead of that? Perhaps physical therapy to get your legs stronger.

By looking at these things in combination, having AI run through all that, we actually have a good chance of identifying patterns early and getting ahead of any future fall or other negative health outcome.

You can listen to our entire conversation here.

Topics

02:10 The Ramen Alarm Moment

03:51 Beyond I’ve Fallen and I Can’t Get Up

06:32 Digital Twins Prevention

10:27 Care Village Apps

11:42 Design Seniors Will Wear

15:46 Future Of Aging Tech

21:24 AI Supports Human Care

27:08 Policy And Personal Advice

31:23 Family Stories And Wrap