Brainstorm AI Singapore 2024: Swallow This! The Transformative Power Of AI-Enabled Micro-Robotics In the Human Body
Torrey SMITH, Co-founder and CEO, Endiatx Moderator: Ellie AUSTIN, FORTUNE
Category
🤖
TechTranscript
00:00So as I said we're talking about micro robots here that are ingested into the body that in itself poses a lot of questions.
00:07It sounds a bit like a science fiction film which is both a source of fascination but also I think for many people probably in
00:14here a source of skepticism. So if we could start off by you explaining to me briefly why you think pillbots is more
00:21effective than the traditional endoscopy technology. Well thank you. Our hope is that with little tiny robots that can swim
00:28around in the body you can actually supplant or augments the existing tech that's out there. I'd like to think that pillbots
00:36could be a little scout to go out and find illness out in the world and then bring that illness into a hospital to treat
00:43properly. OK. OK. So it's it's speeding up the process. It's making it more efficient.
00:49Looking at your Web site the pillbox team it's made up of a lot of engineers. Now some people might expect that to be
00:55doctors scientists on your core team. Can you explain to me why that isn't the case and where the medical expertise comes from in
01:02this process. Well so besides a group of almost 20 deep tech engineers we actually have leading gastroenterologists from all over
01:09the planet that have joined the company and many of them have even stepped up and invested and put their put their names on the
01:16line. And are they part of the team or are they an advisory panel. What's the structure. Well we're very proud to count a vivid
01:23Kumbhari a leading gastroenterologist as one of our co-founders and a very close friend. OK. OK. So some people here in the
01:32room are going. I would imagine to be drawing comparisons to something like Elizabeth Holmes and the Toronto scandal in that
01:40this is an outlandish idea that promises to transform the health care industry. Some might say it's too good to be true. We're
01:49going to see the technology demonstrated in a minute. I mentioned that you're in the process of seeking FDA approval. How else are
01:58you going to legitimize pill ball within the scientific and the medical community to silence those skeptics. Well the key here is
02:06just putting a robot where our mouth is right. If we can create a technology that lets a doctor have a look around inside you to the
02:14point where it's maybe a little bit more inexpensive or a little bit more accessible. And if doctors start to say we'd like to use
02:21this in our actual patients then we're starting to do our job. But until we get to that point where doctors are demanding to use pill
02:28box you know it really is just an R&D project. But we're excited to show it off right here so you can kind of start to judge for
02:34yourself. So it's word of mouth. Obviously that's the FDA component. Is there also going to be peer reviewed research. How does that
02:40side of things work. Actually that's been one of the things that's helping us most of all attract brilliant doctors is to be honest
02:48everyone wants to publish on pillbox. It's an exciting new platform that can open up all all sorts of new research avenues. And so
02:56people are actually realizing that you know joining the team and getting ahold of the robots could actually help them in their their
03:02academic career as well. And are people publishing. I know. Yeah. It hasn't featured hugely in the mainstream press. But if we're
03:08looking at the scholarly academic side of things or kind of what level of publication has there been so far around pillbox. The
03:15biggest publication so far that's coming is that we have officially done our first patients in Auckland New Zealand in our IRB
03:24clinical trials. And we're actually starting to see pillbox do some clinical work. But of course that's just under under under
03:32active progress right now. OK. Once that trial gets complete we'll basically bring all that data together and we can think about
03:39publishing. OK. So nascent. But you're looking to have more data in the future. For sure. My final question before this part of the
03:47session is so pillbox or your your company has raised seven point five million dollars I believe in funding since it was founded in
03:552019. And it's worth noting that one point five million of that dollars cut one point five million of those dollars comes from the
04:02Singapore based Fudge Tech Health Fund seven point five million. It's a good amount. It's not a huge amount. And I imagine that this
04:10is deeply expensive technology to develop. So where's the rest of the money going to come from. Well we're obviously excited to
04:18continue that adventure. You can always do things with new amounts of funding. But the thing that we're really proud of is with the
04:24support of Virg health tech funds here in Singapore and the support of some of the most passionate deep tech engineers in Silicon
04:31Valley and actually beyond. We've been able to make some magic happen using pretty limited resources. But when we show you the robot
04:39today you'll see that it's not perfect. You'll definitely see opportunities for it to improve. And that's kind of where the next
04:46rounds of funding might be able to help us. OK. So you're going to continue pitching right. On that note I know you've got a couple
04:52of slides to show a story to explain in a bit more depth how the robot works. So take it away. Yeah. So let's just briefly look at
04:59a robot and we'll start to ask ourselves you know if we go back in time you know the way that you would get data from someone's
05:06health inside the human body was very difficult. You could cut into the body. You could slide a tube into the body but in a
05:14hospital. And we're asking ourselves could we potentially have microscopic robots travel throughout your body. Not only finding out
05:23what's wrong but maybe actually starting to fix it. So that's the adventure we'd like to go on today. And we'll have to acknowledge that
05:30it's a difficult challenge. But where we're going to begin is the human stomach because in the human stomach we have this unique
05:39opportunity for a robot that is maybe not microscopic. This is the size of your fingertip. But pillbox contains live video three
05:49dimensional motion with pump jets. If you're willing to drink some water you can get some motion and the ability to control it in X Y
05:56and Z. And so the goal here is to basically show there's a part of the body that we can actually send a robot to right now. The
06:07stomach. We can probably do an amazing block of work there. Then where do we go next. And so you'll see right here is kind of where
06:15we are in the current process of miniaturization. I want these to be rice grain size. I want to do brain surgery with it. So looking
06:23inside pillbox you'll see a pretty elegant combination of pump jet motors really high power lithium batteries that we work really
06:30carefully to make safe live video. Right. A lot of this stuff you would see in a drone. But the reason that we're doing these robots is
06:38that they give us the data right. Data is how we understand what's wrong. Data is how we train up a eye. And from here there's a
06:47bright future for this. But you know let's start to talk about what we can actually do with one of these. So OK this this is pill
06:57by. And I need to turn this thing on. And if I'm going to turn this thing on I'm going to need my chief engineer Quentin to come out.
07:05Quentin so up you come. Let's bring this thing to life. OK. So and Tori and Quentin I need you to talk us through what you're doing
07:14here. So Quentin is using an optical signal a bright light to wake pill bot up. It's been sleeping. And so now you'll see pill
07:22bot start to swim around. He just got some water into the thruster lumens so we don't get vapor locked. And what we have here is
07:33basically a swimming little drone. And can I just highlight how Quentin is controlling it. He's using what looks like an Xbox
07:39controller. Yeah. We we find that a lot of younger gastroenterologists really understand the value proposition.
07:47But basically what we're doing here is we're saying let's make a swimming robot. Let's call it a little eyeball. Let's drink some water.
07:55Let's swallow the robot and let's take a procedure that would be done in a hospital after several physical hospital visits. Let's take that
08:03whole adventure and let's digitize it. Let's turn it into a telemedicine call at home. Let's let AI do some of the boring work for us.
08:12So this is about accessibility. How much does a pill cost. So cost of goods can often hover around 35 to 50 bucks. Right. It's a
08:20relatively inexpensive physical article. All of the value is in the hard work that miniaturized it and in the data that comes off of
08:29the robot. But if we can't make it inexpensive to build then there's no way we're going to get it out into the world and actually go
08:35find problems. What do you expect it to cost when it comes to markets in the U.S. market. It'll be a couple of hundred dollars
08:42just like a normal pill camera. But if we're going to take it to Asia and beyond honestly I envision a future where these would
08:50cost maybe 10 or 15 dollars or maybe even less. But you'd have hundreds of millions of people accessing this kind of care. But the
08:58amazing thing is that while we're doing that the health data that would come off a platform like that would be incredibly valuable.
09:05OK. So Tory I believe you are not going to swallow pill bot. I you know honestly we came all this way. I think we might as well
09:15might as well go ahead and swallow. It would be rude not to.
09:20OK. So you're you're gulping down a glass of water to prepare your stomach. I imagine. I need a little fish tank in my tummy. You
09:28need a fish tank in your stomach. OK. And for context Tory told me earlier this is the 45th pill box that he has swallowed in
09:36his lifetime to date. It's it's a lot of fun. We really enjoy this process.
09:46I think for those who can't see I would. Oh wow. More. OK. I would say the pill bot looks like the size of a pretty large
09:54multivitamin. It's it's not small. It's not a small robot. In fact it's probably just about the biggest thing that someone could
10:04ever reasonably swallow. OK. And even maybe half of people might not be able to do it. But the thing that bugs me is that we
10:12transact as much as 55 billion dollars every year in the old way of how we do endoscopies. But that only reaches 1 percent of
10:20stomachs. And so we're we're really not reaching people where we need to reach them. So what do you think. Should we go. We get
10:27this thing. OK. All right. So Quinton is going to hand me this guy and let's let's do it. I'll get it nice and wet. And hopefully
10:39let's get a live video up on the screen here. Yeah. Can we see. Oh here we go.
10:42There we go. OK. So you can see. Oh and I need the antenna a little bit closer to my tummy. There we go. So hopefully you can see me.
10:54You can all see yourselves. And now let's go ahead and swallow pill.
10:59But. All right. So let's have a look at my stomach. I've been starving all day.
11:13I can whip it around a little bit. And then the hope is that we'll sort of break into the bottom of the stomach and you'll kind of see
11:21what's going on.
11:25There we go. So the idea here is we want to make sure that you can quickly and easily give a gastroenterologist a live view inside the
11:37stomach and the ability to move around. OK. I want this to be like the tip of an endoscope but just in virtual form. Wow. OK.
11:45OK. Here we go. We're traveling through your intestine.
11:52Tori if how are you feeling first of all. So sometimes I'll feel it when I swallow it. But honestly once once I swallow it that's OK.
12:00OK. I think we're going to cut the video now because that's enough intestines for one day.
12:05But Tori I've just got a few questions like quickfire questions that I want to ask you as we come to a close. Is there anyone who can't
12:14take pill. But I'm thinking pregnant women and children. Well right now the weird thing is if you're like a child or a cancer
12:23patient you'll have a really delicate GI tract and inserting an endoscope is actually kind of traumatic. And so if we can make it
12:31maybe a little smaller so it would be appropriate for a kid you might actually be opening up treatment avenues for people that are
12:37actually kind of prohibited from them. OK. You said that you mentioned as well that some people can't swallow the pill.
12:44What happens that. Well we have been swallowing camera pills that don't move but still contain many of the components we just saw for
12:53more than 25 years. And during that time we've learned how to do it safely. And we've also learned that maybe one in five maybe one
13:01in 10 people are going to have trouble swallowing a camera pill. So for us even if we make it the same exact size and shape we're
13:10going to be missing out on maybe 10 maybe 20 percent of the population. The thing is though those camera pills because they don't
13:17move. They only address about 1 percent of the population. You can't use them on everyone. And so we're actually talking about
13:24going from addressing 1 percent of people to maybe 80 or 90 percent of people by making it a virtual experience. And worst case
13:34scenario pillbox goes rogue within you. You know what. What are the potential health dangers as a result of that happening.
13:43Because as incredible as this sounds we're putting a small drone into the human body and that must come with risks. Well we had a we
13:49had a pretty amazing chat with FDA back in December and they sat us down. And the thing that was really thrilling was to hear
13:56enthusiasm coming from the biomedical engineers over at FDA from the regulatory bodies from the doctors working with them. And
14:04they said honestly this is tech that everyone is kind of hoping for. And we want to sit down with your team and figure out how do
14:11we do it safely. How do we do it right. And what were they concerned about the FDA when you spoke to them. We were concerned about
14:17the fact that we use lithium batteries and we were curious if FDA would be willing to even entertain the conversation. And they
14:25actually said absolutely. Sit down. Let's go over your engineering plans. Let's see how you test these batteries. Let's see what
14:31happens when a battery fails. Can you contain it. And so far it looks like we're actually on track to being able to turn this into a
14:38safe platform. OK. There's so many more questions I have for Tory. I'm sure you have for Tory as well. But thankfully he will be
14:45sticking around for the next two days. So please do feel free to find him at a coffee break. Flag him down and continue the
14:52conversation because this as I said is quite an outlandish idea. Thank you for demonstrating it to us today. And I look forward to
14:59seeing what happens next. And thank you so much. Thank you.