Self Driving Miss Daisy

00:05 Paul Matzko: Welcome to Building Tomorrow, a show where we envision what the future could look like if we protected the freedom to innovate. The world is getting older. Our median age has been steadily rising for generations. Not only does that mean that we’ll face a serious de-​population crisis by the middle of the 21st century, but the fact that we have fewer kids means that the population in developed countries will skew ever more older. In Japan, the median age has climbed from 29 in 1970, to 48 today, will rise to 55 by 2050. That means that in Japan, there will be as many people aged 56-​and-​up as there are ages one through 55. That’s incredible. And we’re not as far behind as you might think here in the US, with a median age of 38 projected to climb higher, albeit a bit slower. Now an older population will mean new challenges for elder-​care. The cost of nursing homes is already skyrocketing under pressure from the leading edge of retiring baby-​boomers and tight labor markets. There are already too many stories of senior citizens slipping through the cracks of the healthcare system because their families, well, they can’t visit the nursing home every day.

01:13 Paul Matzko: So these early health warning signs are simply missed. As the number of children per family falls even further, the odds that seniors will outlive their one-​or-​two children, it’ll grow. Perhaps as many as a quarter of senior citizens will outlive their single child, along with the growing number of those who never had any children at all will be particularly vulnerable to this problem. But there is hope that we can mitigate some of these problems in the future. If Japan and Korea are a picture of what our grey-​haired future population looks like, they’re also a place where we can catch a glimpse of the future of elder-​care technology. One of the most promising technologies in the field is passive-​health monitoring. These are systems that constantly track a patient’s health rather than relying on periodic visits to the doctor, at best, to catch health problems. I have two interviews that I conducted at TechCrunch Disrupt this year with start-​ups from Japan and South Korea, that use basic home appliances in new ways, including a television that tracks your eye movements and can accurately predict early onset dementia and Alzheimer’s. I’m sitting here with Leland Creswell, the VP… AVP from We Care…

02:26 Leland Creswell: The VP.

02:28 Paul Matzko: The VP. Only one. So extra special here. So, Leland, what does We Care… What’s its vision? What’s its promise?

02:37 Leland Creswell: So We Care, the promise is that we’re able to detect dementia in seniors at extremely early stages when we can actually do think something about it. Yeah, so with my grandfather, he passed away from Alzheimer’s, and one of the main reasons why was because our family didn’t want to accept that he had it. And so by the time he actually entered the healthcare system, he was already at stage 50, where you can’t really do anything anymore anyways, right?

03:10 Paul Matzko: Yeah.

03:10 Leland Creswell: But if we can catch Alzheimer’s and dementia very early, then we’re able to go to the exercises and go to the medications and all of the support that can allow these sorts of patients to actually live very healthy and productive lives much, much longer than if you don’t catch it early. There’s a huge, huge difference. So that is our mission, that’s our big focus.

03:36 Paul Matzko: What’s the tech? Obviously, I can see a brochure here, but our listeners can’t. What’s the technology you’re using to detect?

03:43 Leland Creswell: Yeah, so it’s really interesting. But a lot of the signs that we can use for this detection is based on the way that someone’s eyes are moving, and specifically their gaze.

03:56 Paul Matzko: Interesting.

03:57 Leland Creswell: So we use our deep-​learning AI, along with a stream of video that we take from their home usually, through like a video camera, right? 4K resolution, something like that, hopefully it gives us the best accuracy.

04:12 Paul Matzko: Sure.

04:13 Leland Creswell: But we watch the way that their head moves, and we watch the way that their eyes move, and that gives us enough data to actually begin to predict if they have some form of dementia or even vertigo as well. So the main thing with this actually and the really interesting thing is that when you’re moving your head and when you’re looking at something, there’s two different types of tracking that your eyes do, okay? The first one is called just target tracking, and that’s when you have a moving object, okay? So think about hunter-​gatherers, right? They’re out there and they’re tracking the deer or whatever, jumping around. Your eyes move very smoothly following the target. The second form of eye movement is called saccades, and that’s basically going from one object to another very, very rapidly. So you know you’re walking down the street and you see a pretty lady. Boom, you’re snapping over there, right? It’s a quick movement of your eyes, right? And so with the target-​tracking, ideally what you wanna see is just a smooth tracking, where the eyes are following the line in a very optimal way.

05:19 Leland Creswell: But what we see with vertigo and what we see with dementia and Alzheimer’s, is that the eye will come off of the track that it should be following. It will kind of bob up and down a little bit. And there’s actually… There’s a lot of research about this, about the specific movement patterns that you can use to determine what exactly is going wrong with this person. And then in the second form of eye movement, the saccades…

05:44 Paul Matzko: Yeah.

05:44 Leland Creswell: Oh, that’s always so hard to pronounce that, but in saccades, you want to look at erratic motion. So normally with saccades, your eye jumps directly from one object another, and so on a graph, you’d see that as like a spike in movement and then just flat. But what you see with someone who has dementia, early signs of dementia or vertigo, is that you see the spike, but the spike might be way off target. So they’re looking way to the left of the target that they wanna look at, and then they look back, or you might see a very erratic movement which would be they move to the target and then their eyes are kind of like spasming around the target for a micro-​second or two, and then it comes back.

06:23 Paul Matzko: Yeah. Now is this something that… The way it’s framed in some of your literature is that our senior citizen is watching television, there’s a camera watching their eyes as they watch television, and you’re able to predict based on the eye movements. Is there… Would there be benefit in having a preset program? Like, so there’s a program where you know exactly what’s going on on the screen that should be attracting their attention, and thus, so that kind of provides a baseline. Here’s what a healthy person looks like when they watch this program, here’s what someone who has early dementia or vertigo is seeing and how their eyes track. Is that part of your plans in the long run?

07:02 Leland Creswell: Oh, yes. Exactly. It’s good that you brought that up. So there is a test that we currently use with the TV monitor or with the computer screen where we basically… We push the test to the screen and then it will have basically just a very simple red dot that the senior needs to either follow, and that’s to get the tracking motion, or they’ll have different dots popping up at different locations to get the saccades.

07:29 Paul Matzko: Okay.

07:29 Leland Creswell: So that’s what we are currently doing. In the future though, exactly as you said, for different movies and for different television programs and etcetera, we have already actually found in our research that there is a very, very, very standard average of eye movement across that media, and this is what’s used of course for like advertising online, right? An advertiser will look at eye hotspots, where people normally look on the websites. So this sort of technology exists already and this is also what we want to apply to TV programs and movie programs. It’s not as specific as the red dot test.

08:08 Paul Matzko: Sure.

08:08 Leland Creswell: But it does allow us to actually use the programs that are currently being run on that senior’s television or whatever, and to highly correlate it with their eye movements. It will take us much more data doing it in that way, compared to the dot test…

08:28 Paul Matzko: Sure.

08:29 Leland Creswell: But with a few hours of data and following a program that we have already analyzed, it should be possible though. That’s in progress right now, but it’s really good that you brought that up.

08:38 Paul Matzko: But it’s… It has the advantage also of being less… Not invasive…

08:42 Leland Creswell: Intrusive, right?

08:43 Paul Matzko: Less intrusive.

08:43 Leland Creswell: Yes, very passive. Yes.

08:44 Paul Matzko: They can just do what they’re already doing.

08:46 Leland Creswell: Exactly. Passive… Passive is the main goal of our entire system, because we know that seniors don’t want to be pulling out their smartphone for applications and stuff. A lot of seniors are capable of using apps fine, but some of them, or many of them, I mean my grandma has no idea how to use it and she doesn’t have any interest, right?

09:07 Paul Matzko: Right, right.

09:07 Leland Creswell: Yeah.

09:08 Paul Matzko: And you ask them to download and see a specific program they’re not interested in.

09:11 Leland Creswell: Yes.

09:11 Paul Matzko: You know, I wanna watch my, I don’t know, NCIS.

09:15 Leland Creswell: Yeah, exactly, exactly, right?

09:16 Paul Matzko: I don’t wanna watch…

[overlapping conversation]

09:17 Leland Creswell: Yes, for sure.

09:18 Paul Matzko: So what’s the use case for this? Is this a matter of you imagine somebody contracting with nursing homes to provide this service on a contractual basis for all of their residents or individual families, who’s gonna be using this software?

09:33 Leland Creswell: Yeah, so there’s two main use cases that we’re currently trialling out because, of course, we’re still in the stage where we’re trying to figure out where exactly our market is, right?

09:42 Paul Matzko: Yeah.

09:42 Leland Creswell: The first big one is direct subscriptions to the subscribers or… Sorry, to the seniors themselves. In many cases it would actually be with their family or their guardians…

09:53 Paul Matzko: Someone who has power of attorney, yeah.

09:54 Leland Creswell: You know, the people who are actually, you know… But the idea there is that they pay us a subscription, and then we say, “We will now provide you with protection and early warning about dementia.” And we’re also looking at packaging our dementia protection, our screening, together with other systems that are already being used. For example, like an alarm bell or alarm button. They have something around their neck, they press the button, they get an emergency vehicle right away.

10:23 Paul Matzko: Yeah.

10:23 Leland Creswell: In the Canadian market, which is… I’m more familiar with the Canadian market compared to the American market, but people are paying from 60 to 100 dollars a month just for a button, that if they press it, they’ll get an emergency vehicle right away.

10:34 Paul Matzko: Right.

10:35 Leland Creswell: So, we’re looking at packaging that sort of functionality with our dementia to get into the market.

10:40 Paul Matzko: Yeah.

10:41 Leland Creswell: And then the second main market that we’re looking at is directly with the governments themselves. We have an official relationship with the Seongnam municipal government over in Korea.

10:52 Paul Matzko: Okay.

10:53 Leland Creswell: So they are actually supporting us and paying for us to do our initial beta testing, but in the future we want to go up to governments and say, “Hey, if you deploy these devices to all of your seniors, we can confidently say that we will reduce the likelihood of dementia or we will… ”

11:10 Paul Matzko: Right.

11:11 Leland Creswell: “We will prolong the healthy period of these people who are suffering dementia and keep them out of the healthcare system for, on average, 30% longer or 50% longer.” And the government says to us, “Great, that saves us a lot of money, we’ll pay you on the back end to handle all of our seniors.” So those are the two main markets that we’re looking at.

11:33 Paul Matzko: What’s the time frame? I know you’re still pretty early stage here. You said you have this agreement with the municipal government…

11:40 Leland Creswell: Yes, Seongnam, yes.

11:41 Paul Matzko: In Korea. Are you actually trialling the product in people’s homes…

11:44 Leland Creswell: Yes.

11:44 Paul Matzko: At the moment?

11:45 Leland Creswell: Exactly, that’s exactly what we’re doing right now. There’s a lot of interviews going on with the seniors as they use the devices, as they use the video to see if it’s intrusive in their lives at all, and to show the data to the doctors, of course. And we’re also still trying to get the accuracy to be as high as possible with no false positives, because that’s very important. Though, in our case, the false positive is not as big of a deal as it would be in a medical situation like in a clinic, because false positive, you go to your doctor, “Oh no, you’re okay.” You know, that’s fine, it’s… It’s like an hour or two out of your day, but that’s okay.

12:19 Paul Matzko: Right, right.

12:20 Leland Creswell: Right? But yeah, for our timeline, we are doing the trialling right now in Korea actively. We hope to start a trial in Canada next year quarter two, 2020. And then also we’re going to Singapore to start a trial in 2020, quarter two as well. We’re looking at America as being another place maybe we can start a trial, but we don’t have a lot of traction here, so we’re not really sure if we can actually do something on that early of a timescale, right?

12:49 Paul Matzko: Sure. Yeah, you do it where you have the relationships, and you can always try it here later.

12:54 Leland Creswell: Yeah, exactly. For a start-​up like us, it’s very important for us to try to use our resources efficiently, right? And just trying to throw money out there and get trials going in a bunch of markets, it’s hard.

13:07 Paul Matzko: So South Korea and other parts of East Asia have a particularly large senescence problem… A growing elderly population, as a ratio of the national population.

13:18 Leland Creswell: Oh, for sure, yes.

13:19 Paul Matzko: Now, that’s going to be in the future for many more countries over the next half-​century.

13:23 Leland Creswell: Oh, absolutely, yup.

13:25 Paul Matzko: Can you talk some about what role will that play, and why you guys are situated in South Korea?

13:31 Leland Creswell: Well, South Korea is following Japan. It’s becoming a super-​aged society, right? And so it just… Our CEO had a lot of expertise and experience in the healthcare industry, specifically around using IoT devices in the lives of seniors, and so he just saw that, yeah, the senior market is going to be huge and it… It’s not something that you need any sort of calculation to understand. It’s like, “There’s this many people at this age right now.” And so, from our initial research that we did two years ago, is that not a lot of companies were really moving on this potential. We didn’t see anything coming to market, we didn’t see any market leaders really popping up, and we spoke to the doctors and the senior centers and everyone who was there, and we saw a very slow uptake of technology. So yeah, this looks like something that we could disrupt, especially with a highly passive system.

14:30 Paul Matzko: Yeah.

14:31 Leland Creswell: And that’s why we decided to get into it. And of course, I live in Korea and the CEO is Korean, so it just made sense for us to start in Korea, and it’s kind of… I think, more than planning, it’s kind of happenstance that we are a Korean startup and that we start in that market.

14:46 Paul Matzko: Yeah.

14:47 Leland Creswell: But I mean Canada and America are the same, right? Because Baby Boomers are coming in, they’re entering their silver years already or they’re coming into them, and they’re gonna want to avoid dementia because I know that from back in Korea, I have… So I have a Korean wife and her… So my mother-​in-​law, Korean mother-​in-​law, the only thing that she is worried about is Dementia and Alzheimer’s. She’s not worried about losing her ability to walk or this stuff, she’s worried about forgetting about her family because like… And especially as you get older, I think us as young folks maybe we’re too focused on work or other things, it might not occur to us, but especially for older people, family and connections and the ability to interact with your family is so important and if you lose your memory, that’s it. It’s like dying, right?

15:38 Paul Matzko: Yeah, yeah. Yeah, anyone who’s had a family… A close family member, a grandparent or a parent, who has gone through dementia, Alzheimer’s, one of these disorders, and seeing… It leaves a devastating personal trail through a family. It’s just… It’s very hard. What price are you willing to put on giving some peace of mind that, “Hey, here’s an extra tool that’ll help me detect this kind of disorder early, so that maybe this time it won’t be quite so horrible.” That’s a pretty good sales pitch.

16:12 Leland Creswell: Yeah, I mean, if I could say to someone, “You put this video camera in your house, you pay us a little bit of money each month, and we’re confident that we can catch any sort of problems with potential Alzheimer’s, dementia, early, and we can help you to basically reverse that or to just keep it at a level where you’re okay.” If I can say that to them, I mean, it makes me happy, it makes them happy, it’s just a win-​win-​win, right?

16:39 Paul Matzko: Yeah.

16:40 Leland Creswell: And that’s where you wanna be as a startup. You wanna be where deploying your product will make you money, but it’s also gonna provide real immediate value to the people that you’re working with, right?

16:50 Paul Matzko: We have this funny juxtaposition in tech culture at the moment, where… Or people’s relationship to tech, which on the one hand people love adopting Internet of Things, Smart devices in their home. We like talking to Siri on our smartphones, having Alexa, and having our Amazon’s smart speakers or Google smart speakers in our houses. On the flip side, people are very concerned about surveillance intrusion, the fact that people might be listening to them. So, it’s funny, there’s a bit of a cognitive distance where we both want these things, but we’re also worried about people listening in on us.

17:23 Leland Creswell: Yes.

17:24 Paul Matzko: As a technology that’s predicated on watching people, what’s your response to folks who are concerned about that kind of surveillance?

17:33 Leland Creswell: Well, first off, I completely understand where those people are coming from. I don’t want someone to be peeping on me when I’m just doing stuff at home, right? So I totally understand where they’re coming from. And from our side, we are doing everything we can to make sure that the video itself never leaves the device, like it doesn’t hit the internet, it doesn’t go to our cloud servers where it can be hacked, right? Anyone who says that their cloud system is 100% hack free and they’re just like… They’re totally bullshitting you, right?

18:05 Paul Matzko: Right, yeah.

18:06 Leland Creswell: We’ve seen breaches of Social Security numbers and all this other very, very private data from millions of people from the largest companies. We see it all the time. So anyone who says that data that is out there in a cloud server or whatever is secure is just bullshitting you ’cause it’s not.

18:21 Paul Matzko: Yeah, yeah.

18:23 Leland Creswell: So we’re going from the get-​go, we’re just saying, okay, the data is never going to actually leave that device, the only thing that’s gonna come out of it are the outcomes, like the actual metadata that we’re using to say that this person probably has dementia or probably has vertigo, this sort of thing. And then, as soon as that data has been used by the doctor or by the family or whatnot, then we just get rid of it if it’s not needed, that meta-​data. So it’s not gonna be personally identifiable. And this is something, every time I meet with one of our beta testers or a potential beta tester back in Korea, I say to them, the first thing I say, “I know that having a camera just here in your house, I know the feeling could be that you’re under surveillance.” And so, if there was any other way we could do this more effectively, we would, but the video data is required and we keep it on the device there, and it never goes out of there, and that’s really all we can do. Surveillance, you should be worried about surveillance. There’s a lot of crazy stuff going on but, it’s gonna be required for things like this. We can’t see your eyes, how do we know that you’ve got problems?

19:36 Paul Matzko: Yeah, yeah. And at the end of the day too, if you go to your doctor and have them do a check for early signs of dementia or vertigo, well, they’re also generating medical data, personal data, identifiable data that gets put somewhere, that can be hacked, which has been hacked previously. All of these things generate data.

19:56 Leland Creswell: Yeah, that data is much more personal, the data that the doctor would create when he’s doing these tests for you and then attaching to your file is much, much more identifiable than the data that we’re getting. Because the data we’re getting, we’re only pulling out the metadata, and then later on, it’s attached to your patient portfolio or whatever. And at that point, it’s the doctor’s concern, it’s the hospital’s concern to protect it. From our perspective as a company, we just make sure that the personally identifiable data never actually gets out there. And if someone, the doctor or your family or whatever, someone else wants to take that data and attach it to something, then that’s not us, that’s the doctor or the family or whatever.

20:40 Paul Matzko: What’s the reaction been like from medical professionals, from doctors to… You have someone who’s part of your trial system in South Korea, they’ve got a warning sign for early stage dementia. You then tell them to go to your doctor. What’s doctor’s response been to that situation?

20:58 Leland Creswell: Yeah, yeah, so from the doctors, especially the research doctors, they’re always cool as cucumbers. They come in and we say, “Oh this is what we wanna do, let’s work together.” And they’re like, “Okay, alright. These are the tests we’re doing already. Okay, yeah.” Working with them, I never got the feeling of hype that I get from the tech guys that we work with. When I go to a tech guy or a developer and say, “Hey, we wanna change the world by preventing dementia or helping to prevent it.” They’re like, “Yes! This is a good mission.” [laughter] The doctors are just like, “Yeah. Okay, cool. Okay, alright. Yeah, yeah.” But the reaction is generally positive. This exact technology is not in distribution, it’s not being commercially deployed anywhere in Korea, and at least to my knowledge, in the world right now on a scale. So, most doctors don’t really know about it before we come to them because it’s not deployed. What they have, they have these devices, like these little headsets that are… Basically, they’re stripped down VR headsets.

22:03 Leland Creswell: That are just designed to track eyes very accurately. And these companies that are selling this device, like 5-​6k for the device, for the device that costs 100 bucks to put together. [chuckle] And some software that is just like so, so simple. Literally, they’re just… The software, the only thing that it does is just pulling the data into a graph, and the graph is being used by the doctor, and that is it. We’re talking about 5-​6k for this little software device package. It’s just mind blowing. [chuckle] So, we say to them, “We wanna use just video.” They say, “Oh, that’s interesting,” because they don’t know that it can be done with AI, in some cases. The guys, the doctors who are keeping up on AI and are actually using it in their profession, they say, “Yes. Yeah, okay, this is doable. Yeah, that sounds really interesting, and let’s try getting it out there.” The doctors we talk to who aren’t aware of this data of AI, are kind of skeptical, they’re like, “Hmm, do you have enough resolution? And what about their heads, their head’s moving around. But we need to strap something to them. We need to strap their head in. Where is the straps? [chuckle] Where is the headgear and all this stuff?”

23:11 Paul Matzko: Yeah. Interesting. But you haven’t had any pushback from doctors? So the ones who aren’t particularly familiar with the technology, have you had push… They see it as like competition for what they’re doing in the office. Are there any concerns that…

23:24 Leland Creswell: Not competition.

23:24 Paul Matzko: If you’re doing it, people won’t come to see us or anything like that?

23:28 Leland Creswell: If you spoke to the guys who are selling those headset devices, lots of push back. [chuckle] They’ll say, “Wah, no, no, wah, you need this sort of accuracy, and wah, wah, wah.” But no, from the doctors, they don’t care. If it helps their patient and if it gets them data that they can act on, they’re happy. And especially the doctors that are directly working with the families, like local doctors, just local family doctors, they’re happy to look at the new stuff we’re doing.

23:55 Paul Matzko: Interesting. Now, there’s always been, at least in the US, I don’t know if it’s… How similar it is in Canadian or South Korean markets, there’s… Among medical device startups, there’s always concern about… Well, even using that framing, medical device because once you’re a medical device, then you have to get FDA approval.

24:13 Leland Creswell: Exactly.

24:14 Paul Matzko: Which is a very long, expensive process. So, I don’t know if that’s also true in Canada and South Korea, you can inform us about that. But then, in the US, do you class this as a medical device? Or is this something more like a wearable, like a Fitbit? How do you think of this product in the US context?

24:31 Leland Creswell: Yeah, absolutely not categorized as a medical device. We knew from the get-​go. And especially, like I said, my CEO, he has so much experience in the healthcare system. And Korea, actually has tighter regulations, they’re based off the FDA, but they’re really tight, they’re really tight for getting new devices into the market. So, we knew from the get-​go that we didn’t wanna be classified as an actual clinical device. So instead, we just say, “Hey, we have really great accuracy,” and like I said, false positive, no problem, you waste a little bit of your time but as long as we keep… You come in like one day, false positive, another day, false positive… Okay, no, that’s not good. But having one in a thousand people get a false positive, everyone smiles, no problem for us. But yeah, we don’t want to get into the whole FDA thing because there’s no real need for it.

25:23 Leland Creswell: Preventative… It’s not medicine, but preventative health care, we call it healthy care. Healthy care, it means keeping people healthy before they get sick. It’s all about just watching and processing and analyzing to the best of our ability, as accurately as possible, and then the healthcare system can take over. And so, we really are… I think that this is just something that hasn’t really been available in the past, this sort of… Health care has always really required a very strong doctor… Doctor leading, you know what I mean?

25:58 Paul Matzko: Right, right.

26:00 Leland Creswell: But with our system, because it can be done by yourself or even just as passive, we don’t need to be in the health care system. We can just be there, in your home, protecting you or trying to protect you from dementia by detecting it early or by predicting it early.

26:16 Paul Matzko: So in terms of accuracy, how comparable is it to the device you just described that doctors use with the $5000, whatever, is it comparable in accuracy to detecting signs of early dementia? Or I suppose are you willing to put up with slightly less accuracy on the healthy care side, because like you said, the false positive is not, it’s not as big of a deal as a false negative. So, compare your device, in terms of its efficacy to the devices that doctors currently use to diagnose dementia.

26:47 Leland Creswell: Yeah, sure. So like I said, we’re not getting into the health care side of things.

26:51 Paul Matzko: Right. It’s a opposite of divide that you’re not…

26:53 Leland Creswell: Yeah, exactly, exactly. So, the device is…

26:56 Paul Matzko: But it’s like medical grade even if it’s not medical care?

27:00 Leland Creswell: Well, our opinion is that it’s worldwide cutting-​edge technology, that’s where we’re at in terms of how effective it is, and the level that our researchers are working on. So the quality is equal to or greater than any sort of health care device you see. But the devices that are used in the clinic and in the hospital to say with 100% certainty you have this level of dementia or you have this sort of vertigo, that is a little bit different. So if you look at our accuracy, it is highly accurate, but it is not accurate enough to say based only on that, that you absolutely have this sort of vertigo or you absolutely have this amount or this level of dementia in Alzheimer’s. Instead, we are accurate enough to very confidently say, “You very, very, very, very, very, very likely have dementia and you really need to get your ass down to the doctor’s clinic where you can put yourself through one of these $20,000, $30,000 devices that actually give you the 100%, yes, you do have it.”

28:07 Leland Creswell: Because it’s the same thing with MRI, you can’t go in for an MRI if you have a cold or something, you know what I mean? You don’t… It’s like, “Oh, I feel… Oh, my finger is a little bit sore, I’m gonna go down to the MRI and scan it.” It’s like an entire different level, it’s steps. So we want to be the ultra high quality step that you can just use passively in your home without buying a $30,000 piece of equipment and hiring a doctor to sit there and wait for you to feel a sore thumb or something. [chuckle]

28:36 Paul Matzko: It’s like folks who have… They test their blood sugar. Like there’s much more, closer to 100% accuracy insulin tests with highly expensive medical devices in the hospital, and that’s good, that’s where they should be. But you also want to be able to test frequently with a device that ordinary people can afford…

28:56 Leland Creswell: Yes, exactly.

28:56 Paul Matzko: In the home. And you already can’t recognize that distinction there. Same thing should be true with the dementia. Well, I thank you so much for your time, Leland, this is… It was really enlightening and I really appreciate you talking to us.

[music]

29:09 Paul Matzko: This is why I love covering emerging technology. From this tech to last week’s episode about using smartphones to stop cholera epidemics, it’s a field full of brilliant people working incredibly hard to make the world an incrementally better place. Our next short interview replaces a television with a mirror, a digital concierge powered by AI will answer seniors’ questions about the weather, the news, and so on, while they do their skin care routine or put on their clothes. But it would also take pictures of their faces, checking for sudden changes in skin coloration, which medical researchers have found can be used to check blood pressure non invasively, potentially providing early warning about high blood pressure or even serious heart events. I’m here with Keith Challen, who is the Director for International Business Development for Hanamura. Keith, thank you so much for your time.

30:02 Keith Challen: No, thank you.

30:03 Paul Matzko: So, can you describe in just a few minutes, what it is that Hanamura is trying to do with this mirror concierge system?

30:13 Keith Challen: Mirror concierge is a face and voice interface. It has multiple uses and what we’ve been highlighting today is mirror concierge as a health provider. In Japan, the population is increasing. That’s not true, the population is decreasing and the elderly are increasing the number of people. And so, Prime Minister Abe has an initiative called Society 5.0. And basically, he’s trying to take AI, IOD, Big Data and things like that, using drones, and trying to figure out how to use these to help the aged population. And one of the things that they show on their promotion video is a doctor in a mirror and they show this older woman that showed the granddaughter out the door. In Japan lots of age of families live together, the grandparents with the parents and with the children and grandchildren.

31:26 Keith Challen: And so, the grandmother lets the granddaughter out the door, and then she goes to the computer, and she sits down and she talks to a doctor on the computer, but the doctor is actually an animated character that she’s seeing on the computer. And that’s kind of what we’re doing. We’re taking an avatar and a voice and there are lots of different faces, lots of different voices we can give it. But what it’s really doing, it’s keeping the privacy of the doctor that they’re working with, so that… Literally, the person could be at home in their pajamas probably, and it wouldn’t really matter. What really matters is him seeing the patient and seeing their face color and with the camera, they’re actually able to open their mouth and see down their throat a little bit. And then this particular device which was made by a company or a university, medical university in Japan, is able to check your heart rate, your pulse, your blood pressure and stress level.

32:29 Paul Matzko: All just from video, it can check.

32:30 Keith Challen: Just from the video, yes, just from the camera. And it’s a new technology that they developed and you can see here, it is looking at different colors in the face and how the colors change, it’s able to understand what their blood pressure is and what their heart rate is.

32:51 Paul Matzko: So now do they take… I’ve done the demonstration here where you stand, it scans you for like… It watches you for 10 seconds.

33:00 Keith Challen: Ten seconds.

33:00 Paul Matzko: And then, it watches from a baseline how your over time, how your skin color changes.

33:08 Keith Challen: My understanding, from the technology, from the explanation I receive, it’s basically taking quick photos of your face and that the colors in your face are changing very quickly. And that change of the coloring in your face gives them the ability to be able to understand your blood pressure. And then…

33:29 Paul Matzko: And now, this is being made in partnership, you said, with a medical university?

33:35 Keith Challen: It’s a medical university in Japan. And basically, they came to Hanamura and they ask us, “Okay, we have this technology, how can we use it in real life?” And that’s how we came about being involved in this.

33:51 Paul Matzko: So, the idea is you’re a child, you have an elderly parent, who in the US context, it’s much less likely that elderly parents live with their children anymore. So you’ve got a parent who’s maybe in an assisted living facility, but you want to be able to have some sense of like, “Okay, how are they doing? What’s a baseline rate for their… ” Not medical information, per se, I suppose, but pre-​medical information. You want to keep kinda tabs, make sure they’re okay. How do you go about marketing this system to them potentially?

34:27 Keith Challen: Well, ideally, the marketing would be in several areas. Actually, one person came by earlier today and they were talking about the insurance industry and that this could actually be a very large, cost-​cutting application for the insurance industries, instead of them having to actually go in and see a doctor on a weekly basis, depending on their condition. They can just go in front of the mirror and do it, and the data would be sent in to the doctor. And so, the insurance company would cut their costs because of that. It could have many implications.

35:07 Paul Matzko: You have the potential advantage too, I imagine, for… Right now, there’s a recommended schedule for the elderly to go see or have larger care facilities, having a nurse come and check vital information every so often, every week, every month. But in theory, with this technology, you can do that or at least get some of that information constantly every day. So you’re able to detect things quicker and having to wait for the week to notice, “Oh, for the last week, your blood pressure’s been spiking.” We can address that more rapidly in theory. So I imagine it could be good for care as well. Better care and cost savings on both ends.

35:46 Keith Challen: Yeah. As you can see here, they actually call it the daily checkup. And basically, as you go… In the morning, after your bath or shower and you’re brushing your hair, she’s getting her face scanned and she’s also standing on top of a scale here which has fed into the database and all of that data is stored. It could be stored locally on an app that you can have that they have developed now for the phone and for an iPad, or it could be stored in the cloud, or it could actually be set up where it would be sent to her doctor, his or her doctor.

36:21 Paul Matzko: Yeah. What’s the current prototype roll-​out plan in Japan? Are there some facilities that are trialling it right now, or is it you’re still hoping that you get into homes or in the facilities?

36:34 Keith Challen: I would say, right now, this particular use of doctor concierge is being used in the model homes. They call them model homes in Japan. Each of the major companies, they have these areas where their prototypes of their homes are setup.

36:50 Paul Matzko: So, the home of the future kind of thing, okay.

36:52 Keith Challen: Yeah, and a lot of the companies have come to Hanamura asking to have these put into their bathrooms. And, yes.

37:00 Paul Matzko: I see. Well, Keith, thank you so much for your time. I wish you all the best and our listeners I think have really enjoyed hearing about the product, so thank you.

37:08 Keith Challen: Thank you.

[music]

37:12 Paul Matzko: For me, the most exciting thing about this tech is that for every application I can imagine, there are dozens I can’t even begin to think of. Imagine a future in which your smart clothing and accessories, which can get power themselves from the friction generated by the fibers rubbing against each other while you walk, is able to track your temperature, eye movement, blood pressure, really just get a constant real-​time picture of your health. An AI that monitors changes in your health baseline for concerning signs and kicks those to your physician for consultation. We could stop or mitigate so many heart attacks, strokes, catch cancer earlier, prevent repetitive use injuries for athletes and workers, and so much more, but… And this is a million-​dollar coordinating conjunction, even as we develop that technology, that I just laid out should be possible in our lifetimes, there’s a huge barrier that prevents that from happening.

38:02 Paul Matzko: The food and drug administration. The FDA has incredibly onerous regulations on medical devices that stymie innovation and ultimately cause unnecessary suffering and even deaths on the massive scale. It delays innovation. Everything from breast pumps to insulin pumps. The future I just laid out could never happen because the FDA has erected the high wall between the rapidly innovating world of wearables and other consumer devices, like I just described in this episode, smartwatches, Fitbits and the like. A high wall between those wearables and consumer applications and medical devices.

38:38 Paul Matzko: Every startup I spoke to avoids coming even close to the medical device field like a plague just because of the regulatory morass and the powerful incumbent interest involved. And as a result there’s very little interaction between all your wearables, the data they generate, and the healthcare system. That means that the data can’t make its way through this non-​porous artificially erected barrier. Lifesaving and life improving innovations are being delayed or limited for woefully inadequate reasons. You can go back and listen to episode 32 if you wanna know more about how that world works or doesn’t, as the case may be, but know that we can do better, that there are promising innovators working on technologies that will improve your and my health prospects in the future, if only we let them. As always, but especially today, until next week, be well.

[music]

39:30 Paul Matzko: Thanks for listening. Building Tomorrow is produced by Tess Terrible. If you enjoy Building Tomorrow, please subscribe to us on iTunes or wherever you get your podcasts. If you’d like to learn more about libertarianism, find this on the web at www​.lib​er​tar​i​an​ism​.org.