[00:01.680 --> 00:03.740]  My name is Michael Dirkus.
[00:03.940 --> 00:08.720]  I'm going to be talking to you about how to essentially build your own glucose monitor.
[00:09.020 --> 00:11.560]  Now, I'm 17.
[00:11.580 --> 00:18.640]  I'm not exactly some sort of biomedical engineer, so you're not really going to be getting any, like, secret information here.
[00:18.860 --> 00:29.240]  But, as I'll show you in a little bit, just the process of engineering a glucose monitor and the basics of how it works are really interesting.
[00:29.780 --> 00:36.960]  As a bit of a start, I'd like to tell you all why I did this research.
[00:37.300 --> 00:49.200]  At my high school, there's a club called SkillsUSA, and my friend David over here told me around November of last year that I should join it and work with him on a project.
[00:49.540 --> 00:54.280]  I agreed, and we decided to research how to build our own glucose monitor.
[00:57.240 --> 01:01.140]  The specific competition in SkillsUSA that we did is engineering design.
[01:01.400 --> 01:14.040]  So, although it wasn't a functional prototype, we were able to do tons of research on it and actually get some form of functionality, at least, with the sensor itself.
[01:16.040 --> 01:23.800]  With that being said, today we're going to be digging into almost four paradigms of building a glucose sensor.
[01:23.960 --> 01:27.600]  First of all, why should we even care about this?
[01:27.880 --> 01:30.700]  Which is almost the economical standpoint of it.
[01:30.760 --> 01:32.200]  Glucose monitors are expensive.
[01:32.400 --> 01:33.580]  We need alternatives to them.
[01:33.680 --> 01:34.480]  We need improvement.
[01:35.160 --> 01:37.360]  Secondly, the chemistry and biology.
[01:37.620 --> 01:38.940]  What is the sensor made of?
[01:39.100 --> 01:39.820]  How does it work?
[01:39.820 --> 01:42.100]  Next, the hardware.
[01:42.700 --> 01:48.140]  Essentially, what are the design goals that we should be looking for when we're building a glucose monitor?
[01:48.380 --> 01:51.020]  What communications protocols should it use?
[01:51.220 --> 01:53.760]  What power mechanisms should it use?
[01:54.120 --> 02:01.840]  And lastly, just a quick discussion on the physics of the glucose sensor itself and just conductivity sensors in general.
[02:03.080 --> 02:12.340]  So, to start out, glucose monitors are, of course, extremely important, especially for the 38 million people in America that have diabetes.
[02:12.720 --> 02:17.300]  That's 11.6% of the population and it's only expected to grow over time.
[02:18.700 --> 02:28.080]  Just as a definition of diabetes, generally, diabetics are not able to regulate or have trouble regulating, generally, their insulin production.
[02:28.880 --> 02:35.060]  Insulin is a hormone that boosts absorption of glucose from the body into its cells.
[02:35.260 --> 02:42.620]  So, essentially, think of it as an intermediary between, say, all of the functions of your body, like movement and sugar.
[02:42.900 --> 02:46.280]  At the end of the day, if you need to eat, you need insulin.
[02:48.100 --> 02:55.580]  Diabetics must maintain a stable blood sugar level throughout the day, which can be done either via an injection or an insulin pump.
[02:55.580 --> 03:03.260]  If it's not properly maintained, then they can have trouble in their heart, their kidneys, or their nerves.
[03:03.460 --> 03:05.780]  That's something that you definitely do not want to deal with.
[03:05.940 --> 03:13.880]  So, having a glucose monitor to figure out what your blood sugar level is at all times is extremely important.
[03:16.160 --> 03:26.500]  In terms of how I can figure out what my blood sugar is, well, like I just said, you can use a continuous glucose monitor, which will tell you up-to-date information and even give you trends.
[03:26.820 --> 03:34.580]  So, if I eat lunch, it will pick up on that and will actually tell me that my glucose level is going to spike in the coming hours.
[03:34.580 --> 03:43.360]  On the other hand, if you don't have a glucose monitor and you don't have a finger prick monitor either, what you're going to see is nothing.
[03:43.780 --> 03:49.260]  In severe cases, people who can't monitor their glucose may end up losing their vision.
[03:49.620 --> 03:54.300]  It's a very serious consequence of not monitoring your glucose.
[03:54.420 --> 03:59.860]  And it does happen to people, especially those in third world countries that can't afford glucose monitoring.
[03:59.860 --> 04:02.700]  So, overall, we need glucose monitors.
[04:04.100 --> 04:08.220]  Next, I'd just like to point your attention to two different types of glucose monitors.
[04:08.840 --> 04:22.200]  Although the continuous glucose monitor is gaining relevance, especially nowadays, lots of people still use the traditional finger prick sensor, which is more affordable, but at the same time less convenient as you have to prick your finger every single time you want to see your glucose level.
[04:22.420 --> 04:30.500]  It also doesn't really work well in terms of seeing your trends of glucose, which a continuous glucose monitor can do.
[04:31.080 --> 04:33.860]  Continuous glucose monitors are very expensive.
[04:34.280 --> 04:35.120]  Very expensive.
[04:35.820 --> 04:39.280]  But at the same time, for a lot of people, they're worth it.
[04:39.580 --> 04:41.220]  Especially if they're covered by insurance.
[04:41.460 --> 04:46.440]  Because you can view your glucose level effectively at any time.
[04:46.840 --> 04:53.460]  Modern glucose monitors will update about every five minutes, which you're not going to be pricking your finger every five minutes.
[04:53.460 --> 04:59.700]  So you get more data to work with, which means a better view of your overall diabetes.
[05:00.420 --> 05:07.020]  At the end of the day, though, what we'll see in the coming slides is that these technologies are essentially the same, just in different packages.
[05:07.260 --> 05:10.740]  So the price difference isn't really coming in the form of the sensor.
[05:12.600 --> 05:20.040]  There's been a lot of talk lately about building your own glucose monitor, especially in the past decade.
[05:21.300 --> 05:30.980]  Not only in the hacking community, but also in places like BBC or the Wall Street Journal or Wired.
[05:30.980 --> 05:35.120]  It's really a popular subject because it represents bodily autonomy.
[05:35.680 --> 05:39.180]  These devices are permanently connected to us, in a sense.
[05:39.680 --> 05:44.200]  And so being able to control them is our bodily autonomy.
[05:44.200 --> 05:48.160]  It's attached to you in a literal and a metaphorical sense.
[05:49.320 --> 05:50.800]  So why is this happening?
[05:50.920 --> 05:52.900]  Why is there so much reporting on this right now?
[05:53.160 --> 05:56.040]  Well, first of all, glucose monitors are really expensive.
[05:56.220 --> 06:15.780]  If you are unfortunate enough to be in the middle group of people who aren't covered by Medicaid and also aren't covered by an employer-sponsored or self-paid health insurance, you're going to be paying $15.20, not per week, not per month, but every single day that you're alive.
[06:15.780 --> 06:23.320]  That's $5,500 per year spent just on a glucose monitor, something that you effectively need as a diabetic.
[06:23.880 --> 06:38.260]  6.6% of adults with diabetes don't have any health coverage, which means that there's going to be... there's 1.56 million people who either are covered with Medicaid or actually have to pay that $5,500 per year amount.
[06:38.980 --> 06:46.140]  Now, 15.3% of diabetic adults were insured in 2009, so we're improving, but that's still a huge, huge number.
[06:46.460 --> 06:49.220]  And it's not only a problem if you're uninsured.
[06:49.240 --> 06:52.080]  You're paying for it either indirectly or directly.
[06:52.080 --> 07:00.180]  You're paying through it for your tax dollars with Medicaid, or you're paying through it with higher premiums for your insurance provider.
[07:00.180 --> 07:04.680]  So no matter what, your glucose monitor is expensive.
[07:06.040 --> 07:11.420]  And with modern technology, it's only going to keep getting more expensive, in a sense.
[07:11.980 --> 07:14.020]  And it's not like you can buy a used glucose monitor.
[07:14.360 --> 07:17.000]  You don't want to be sharing needles with other people.
[07:17.560 --> 07:21.400]  And, I mean, honestly, are you really going to do that anyway?
[07:21.400 --> 07:24.680]  But the mold.
[07:26.400 --> 07:31.580]  I know this seems a little random, but glucose monitors kind of have to do with mold.
[07:32.020 --> 07:36.520]  This is getting into the biology side of things, in a way.
[07:37.060 --> 07:43.320]  But every single glucose monitor uses mold to measure your blood sugar.
[07:43.940 --> 07:45.280]  It's not quite mold, though.
[07:45.520 --> 07:46.080]  Don't worry.
[07:47.320 --> 07:52.780]  What they do is take an enzyme of a very common mold spore called aspergillus.
[07:54.560 --> 07:56.240]  It's literally everywhere.
[07:56.240 --> 07:57.880]  You probably breathed it this week.
[07:58.160 --> 08:08.740]  It's not particularly fun if you're immunocompromised, but you're not going to be having problems if you are, you know, the average human being.
[08:11.140 --> 08:20.760]  If you are immunocompromised, though, you're going to be experiencing symptoms like aspergilloma, which is a ball of mold in the lungs, not too fun, or aspergillosis.
[08:20.900 --> 08:21.560]  Those are serious.
[08:21.680 --> 08:22.980]  They require hospitalization.
[08:23.180 --> 08:29.660]  So why are we using this crazy mold in diabetes monitors, in medical devices that are literally attached to you?
[08:29.660 --> 08:46.420]  Well, the reason why is because it actually has an important chemical reaction that can turn glucose into hydrogen peroxide.
[08:46.680 --> 08:50.620]  It's also used for a bunch of other things, like food production.
[08:50.620 --> 08:53.400]  It's used in cattle as a de-bloating agent.
[08:53.600 --> 08:54.620]  I know that sounds weird.
[08:54.840 --> 09:01.620]  In some countries, especially China, they'll use it instead of antibiotics.
[09:02.600 --> 09:09.160]  Antibiotics can serve as a bit of an anti-bloating agent too, but they cause all sorts of problems in the cattle feed industry.
[09:09.500 --> 09:10.340]  Our U.S.
[09:10.440 --> 09:13.240]  cattle industry has dealt with this with even more antibiotics.
[09:13.240 --> 09:17.080]  In other countries, they can't do that, so they have to use this.
[09:17.080 --> 09:20.600]  It's also used as an alternative for yeast.
[09:20.940 --> 09:26.260]  So if you ate bread recently, this could have been in there in one way or another.
[09:27.400 --> 09:29.600]  We had to buy this from China.
[09:29.600 --> 09:34.180]  It's very hard to get your hands on, especially as a 17-year-old student.
[09:35.680 --> 09:37.580]  They usually regulate these things.
[09:37.800 --> 09:43.380]  But we were able to find it on AliExpress, of all places, as always.
[09:45.860 --> 09:46.680]  We got it.
[09:46.880 --> 09:49.800]  It took like two months to arrive, as usual.
[09:50.320 --> 09:52.260]  It takes like one day to get to the U.S.
[09:52.360 --> 09:54.100]  Then it just sits in California for a month.
[09:54.380 --> 09:55.660]  You guys know how it goes.
[09:57.100 --> 10:01.620]  And the first time, keyword the first time, they got all the way to our high school.
[10:01.980 --> 10:03.100]  And then they lost it.
[10:04.440 --> 10:06.280]  So we had to order it again.
[10:07.100 --> 10:09.000]  And it was getting really close.
[10:09.000 --> 10:10.060]  It was getting tense.
[10:10.760 --> 10:15.320]  But it ended up coming literally like, I think, a week before our competition.
[10:16.120 --> 10:17.880]  It all worked out at the end.
[10:18.080 --> 10:20.380]  But I'm just happy that it did.
[10:20.860 --> 10:23.260]  Back to the whole chemistry of the situation.
[10:24.340 --> 10:26.400]  I don't know how many of you are chemists.
[10:26.660 --> 10:27.840]  I certainly am not.
[10:28.180 --> 10:29.380]  This looks like gibberish to me.
[10:29.940 --> 10:37.400]  So I made it a bit easier to understand for all of you guys that don't want to deal with that.
[10:37.400 --> 10:42.800]  So essentially, glucose oxidase is the enzyme that we're using.
[10:43.140 --> 10:47.080]  That's what mold... that's what the mold is producing as an enzyme.
[10:47.780 --> 10:52.080]  We take... what a sensor will do is it'll have glucose oxidase in it.
[10:52.440 --> 10:56.660]  Literally in the sensor, in the needle that is in your bloodstream.
[10:57.440 --> 10:59.880]  It'll take a sample of glucose.
[11:00.520 --> 11:02.880]  It'll take oxygen from the ambient air.
[11:03.400 --> 11:06.780]  It'll mix that with the glucose oxidase and it'll produce hydrogen peroxide.
[11:07.400 --> 11:10.880]  So now the question is, why do I need hydrogen peroxide?
[11:11.360 --> 11:12.480]  Seems kind of random, right?
[11:12.760 --> 11:13.400]  That's bleach.
[11:14.440 --> 11:16.640]  We don't need a bleach factory attached to our arm.
[11:18.320 --> 11:21.540]  Well, we can measure the conductivity of hydrogen peroxide.
[11:22.000 --> 11:27.240]  We can't measure the conductivity of sugar in your blood, but we can measure the conductivity of hydrogen peroxide.
[11:27.400 --> 11:31.360]  What we can actually do with that, then, is just make a conductivity sensor.
[11:32.160 --> 11:44.580]  All we have to do is measure the resistance, essentially, of the hydrogen peroxide, and then make a model between how much hydrogen peroxide there is and how much blood sugar you have.
[11:44.780 --> 11:48.400]  And every commercial glucose monitor will do this in some respect.
[11:50.140 --> 11:53.060]  This kind of boils down to two steps, then.
[11:53.240 --> 11:59.000]  We need to build a conductivity sensor, and we need to make a model for converting conductivity into glucose concentration.
[12:00.400 --> 12:03.120]  So now we get to building the actual conductivity sensor.
[12:04.900 --> 12:08.640]  Just a bit on how conductivity measurement works quickly.
[12:08.860 --> 12:13.060]  This is something that you would do, say, if you wanted to test water quality.
[12:15.580 --> 12:19.040]  All of you know Ohm's Law from your high school physics class.
[12:19.160 --> 12:20.720]  Voltage equals current times resistance.
[12:21.040 --> 12:29.580]  And we can just use that simple principle to figure out what the conductivity between us in a solution is.
[12:29.580 --> 12:32.620]  We induce a known voltage across two electrodes.
[12:33.060 --> 12:43.280]  Those two electrodes, as I'll show you in a bit, will be part of the needle itself in a continuous glucose monitor, or will be part of a test strip in a fingerprint monitor.
[12:44.400 --> 12:55.780]  After you induce a voltage between the two electrodes, you measure the resulting current and get resistance.
[12:55.780 --> 13:00.960]  From resistance, we can figure out conductivity as they're inversely proportionate.
[13:01.320 --> 13:18.960]  You'd essentially have to model it with different glucose concentrations to figure out the relationship between glucose concentration and the amount of hydrogen peroxide that is being produced.
[13:18.960 --> 13:23.480]  And that would require calibration.
[13:23.800 --> 13:26.620]  That's why some glucose monitors require calibration in the beginning.
[13:26.860 --> 13:28.060]  Each person is different.
[13:28.280 --> 13:34.980]  Each person is going to have a different hydrogen peroxide to glucose concentration curve, in a way.
[13:36.640 --> 13:42.420]  It's also important, when people are developing these, to minimize the resistance of the electrodes.
[13:42.420 --> 13:44.340]  Think of it this way.
[13:44.800 --> 13:52.020]  If I'm measuring the hydrogen peroxide, I can't have my test leads affecting the measurement.
[13:52.180 --> 14:01.100]  I can't have the resistance of them adding to the resistance of the hydrogen peroxide, creating an inadequate reading.
[14:01.680 --> 14:12.480]  An inaccurate reading is not only bad for us, the glucose manufacturer designer, hypothetically, it's also bad for the actual person using the glucose monitor.
[14:12.700 --> 14:23.500]  They'll see an inaccurate glucose monitor reading, and either pump too much insulin into their body, or not pump enough insulin into their body, either of which is going to result in serious side effects.
[14:26.120 --> 14:44.660]  Now, this looks a little crazy, but around winter of last year, I had a project, slightly different but similar principle, for a science team competition, where we were measuring the conductivity of salt.
[14:45.160 --> 14:58.200]  And it actually worked out really well in the end, because after hours of preparation and failure, literally, it didn't work until the very last day.
[14:58.200 --> 15:02.600]  But it worked in the end, which is good.
[15:03.800 --> 15:19.260]  Not only were we able to prove that a conductivity sensor for this specific instance would work, but also that us, as a SkillsUSA team, might be able to build a real glucose monitor, or at least the sensor.
[15:20.520 --> 15:27.100]  So from that, let's look at what a real sensor looks like, not what my crazy contraption is.
[15:27.100 --> 15:30.000]  On the left, we have the standard finger prick monitor.
[15:30.240 --> 15:32.820]  And on the right, we have a continuous monitor.
[15:33.520 --> 15:37.940]  Now, these three on the right are actually part of the needle apparatus.
[15:38.120 --> 15:54.200]  So this would be in your blood, in your bloodstream, taking samples and measuring the amount of, well, measuring the amount of hydrogen peroxide because of that glucose oxidase I was talking about before, that mold enzyme.
[15:55.120 --> 16:02.520]  If I take your attention to number B on the right, you can actually see the two electrodes.
[16:02.820 --> 16:10.540]  The silver one, that's kind of like in the middle here, adds one electrode and then the coil around it is another.
[16:12.120 --> 16:17.240]  And secondly, in C on the bottom, you can see where the glucose oxidase would be stored.
[16:18.140 --> 16:26.780]  Typically, there'll be some sort of vinyl or even a thin layer of foam that'll work to hold the glucose oxidase in while it's being measured.
[16:28.000 --> 16:31.420]  And then on the left, we have the fingerprint monitor, which is quite similar.
[16:31.720 --> 16:33.720]  We also have an electrode.
[16:34.000 --> 16:39.560]  This time it's called the working electrode because we also have separate electrodes to make sure you don't put too much blood in there.
[16:40.360 --> 16:44.020]  And we have the enzyme on a little strip right there.
[16:45.260 --> 16:49.180]  Some of these monitors will also have a separate chemical added to them.
[16:49.280 --> 16:53.760]  That'll change the color of the test strip based on the amount of glucose.
[16:54.000 --> 16:55.540]  Some will, some don't.
[16:55.540 --> 16:58.920]  So if that's the case, it'll be mixed in with that enzyme there.
[17:05.060 --> 17:10.200]  Next, we decided, hey, why don't we actually try to make a sensor that will measure glucose conductivity?
[17:11.100 --> 17:16.540]  I know this looks really weird and it's way too big to actually use as a glucose monitor.
[17:16.840 --> 17:28.340]  But if I wanted to build a real glucose monitor sensor, I'd have to either have really, really small hands or some sort of nanofabrication lab, which a high school isn't going to have.
[17:28.340 --> 17:31.080]  Maybe a really nice high school, but, you know.
[17:33.340 --> 17:37.060]  The other solution was to compromise, which is what we ended up doing.
[17:37.060 --> 17:48.600]  We made a big glucose sensor and tested it in a beaker with sugar water, which acts as a similar medium to blood, just without all the other parts.
[17:50.440 --> 17:58.660]  The other important thing with this, bit of a cost-cutting measure, we used copper instead of other electrode materials like silver or gold.
[17:58.880 --> 18:09.920]  You know how I told you before that you want to use very low resistance material for your electrodes because otherwise it'll interfere with the accuracy of the readings.
[18:10.600 --> 18:12.160]  Copper is okay.
[18:12.160 --> 18:13.500]  You can get better.
[18:13.740 --> 18:15.540]  It's pretty close to good.
[18:16.020 --> 18:17.300]  So that's what we went with.
[18:17.420 --> 18:19.420]  And it's certainly a lot cheaper than gold.
[18:19.980 --> 18:21.460]  Funny story about this too.
[18:21.720 --> 18:43.680]  Around 7 in the morning, my friend David and I, he was the second from left on the slide with Skills USA, got in my car, drove to his house, and then we went all the way up to downtown Chicago to go to McMaster Car, which sells all of these random things like nuts and bolts and copper bars and foam.
[18:43.820 --> 19:03.040]  And we just got a bunch of stuff, went back to school the next day, got to cutting the copper bars, and then went into my chemistry teacher, Mr. Kane's room, with a box full of glucose oxidase that smells horrible and a bunch of cut-up pieces of copper bar while he was about to leave,
[19:03.080 --> 19:06.800]  and told him, hey, can we stay in your room for a second?
[19:06.960 --> 19:08.140]  It was not a second.
[19:08.560 --> 19:10.120]  But we ended up making it.
[19:11.820 --> 19:12.840]  And it kind of worked.
[19:13.260 --> 19:14.240]  I'll get to that later.
[19:15.560 --> 19:19.580]  If you look in the top right, I have kind of like a sandwich diagram of how it worked.
[19:19.780 --> 19:23.260]  I mean, it's pretty obvious, but just this picture isn't very clear.
[19:23.480 --> 19:27.120]  So you have your one electrode and your membrane in the middle.
[19:27.120 --> 19:31.300]  This is just literally foam that we found at David's house.
[19:31.800 --> 19:35.500]  And it's got glucose oxidase essentially in it.
[19:35.640 --> 19:40.780]  We just soak it in glucose oxidase in water, take it out, let it dry for a little bit.
[19:41.480 --> 19:43.600]  It's not a very scientific test, by the way.
[19:43.700 --> 19:47.380]  Like, please don't repeat this if you're building an actual glucose monitor.
[19:49.880 --> 19:50.600]  So yeah.
[19:50.960 --> 20:01.120]  We dip the whole thing after we get the glucose oxidase mixed in there into a beaker with sugar water, attach some probes onto it, and then measure away.
[20:01.380 --> 20:02.500]  And measure we did.
[20:02.980 --> 20:06.140]  We only measured four times because it was right before we had to leave.
[20:06.560 --> 20:08.840]  So again, take this with a grain of salt.
[20:09.560 --> 20:14.500]  But we actually did get 0.98R value, strong-ish correlation.
[20:15.060 --> 20:17.560]  Again, with only four data points, you can't really tell.
[20:17.560 --> 20:24.680]  But hey, it was literally two pieces of copper with a piece of foam in the middle, super glued together.
[20:24.860 --> 20:25.680]  And it worked.
[20:25.920 --> 20:29.260]  So that must mean the principles are true.
[20:32.100 --> 20:35.200]  Next up, we're going to get more abstract with the hardware.
[20:36.040 --> 20:40.340]  This isn't necessarily discussing, like, if I were to tear down a Dexcom, what would be in it.
[20:40.420 --> 20:46.220]  But more just, like, what is a manufacturer looking for when they develop a glucose monitor?
[20:47.440 --> 20:50.880]  First of all, as with any engineering project, you should set some goals.
[20:51.540 --> 20:56.300]  One of the big, most important goals with a glucose monitor is low power consumption.
[20:56.760 --> 21:01.020]  Nowadays, there are glucose monitors that literally have the thickness of a quarter.
[21:01.660 --> 21:03.720]  And they still have batteries in them.
[21:05.440 --> 21:10.500]  So, being able to use as little power as possible is very important.
[21:10.600 --> 21:12.240]  We want something that lasts a long time.
[21:12.360 --> 21:14.200]  We don't want to have to replace it every week.
[21:14.860 --> 21:20.520]  At the same time, and a bit against this, we want high accuracy and high measurement frequency.
[21:20.660 --> 21:22.040]  Those are going to use more power.
[21:22.660 --> 21:28.160]  Dexcom will take a measurement around every five minutes.
[21:28.160 --> 21:31.660]  Each time it's measuring is time that it's taking up power.
[21:31.960 --> 21:42.460]  So, we need to be able to balance the increased power consumption of more measurements with the benefits of more data or analytics.
[21:44.220 --> 21:49.360]  We also need it to be able to interact with some sort of separate device.
[21:49.520 --> 21:54.420]  I don't think you guys want a screen attached to your arm all the time, even if it's really small.
[21:54.420 --> 22:01.080]  So, you need something to actually be able to read or even chart what your glucose level is.
[22:02.000 --> 22:05.040]  Before, that would typically be a standalone device.
[22:06.800 --> 22:23.160]  Nowadays, it's starting to be more common to have, say, a mobile phone or, I think even in some rare instances, a USB dongle that you'd connect to your computer to just view what your glucose monitor is reading, reading, view the history.
[22:23.740 --> 22:31.960]  Also, some glucose monitors will upload data to the cloud to view your trend, but in our example, that doesn't exist.
[22:33.380 --> 22:36.880]  And lastly, like I was saying before, we need a small footprint.
[22:37.140 --> 22:39.920]  You can't have a big, giant box on your arm.
[22:42.300 --> 22:43.300]  Power sources.
[22:44.700 --> 22:50.600]  There are two options, really only one, but two options for powering a glucose monitor.
[22:50.860 --> 22:55.280]  The main one, the one that 99% of glucose monitors are going to use, is just a battery.
[22:56.120 --> 23:02.360]  It's the current standard, but it needs to be replaced somewhat frequently, as it is a battery.
[23:02.360 --> 23:06.520]  So it's not going to be able to be recharged by anything.
[23:07.040 --> 23:09.340]  And inevitably, you're just going to need to get a new one.
[23:09.560 --> 23:19.420]  On the other hand, and this is what we did a little research on as well in our SkillsUSA project, is how can we apply energy harvesting to a glucose monitor?
[23:21.460 --> 23:26.100]  You can use, say, movement or heat to produce electricity.
[23:26.100 --> 23:37.520]  It won't be a lot of electricity, but it'll be a really... enough to do some sporadic measurements that you could then use to have essentially like a self-powered glucose monitor in a way.
[23:38.740 --> 23:43.080]  That is obviously more expensive, because you're going to have more parts to deal with.
[23:43.160 --> 23:44.900]  You're going to have a higher failure rate.
[23:45.800 --> 23:49.480]  And also, you're not going to be getting nearly as much power as you would with the battery.
[23:49.680 --> 23:53.380]  So you're going to be taking a hit in terms of how many measurements you can take.
[23:53.380 --> 24:03.680]  I don't know about you, I'd rather have accurate and fast measurements than essentially a 20-year long glucose monitor.
[24:05.300 --> 24:08.140]  The footprint for the device is also going to be huge.
[24:08.780 --> 24:16.240]  Most microcontrollers will operate, you know, around, you know, one to three volts-ish for this sort of device.
[24:16.620 --> 24:21.420]  And a glucose monitor is going to require that even for the sensor.
[24:21.420 --> 24:29.800]  So you need to step up the voltage from what is literally like in the microwatts to...
[24:31.180 --> 24:31.700]  Sorry.
[24:32.380 --> 24:37.120]  You need to step up the voltage from what's in the milliwatts or even the microwatts.
[24:37.760 --> 24:39.160]  Why do I keep calling them watts?
[24:39.540 --> 24:45.540]  Anyone who knows physics 1 or E&M, they should be really mad at me right now.
[24:47.580 --> 24:53.020]  But with voltage, millivolts or microvolts need to turn into volts.
[24:53.420 --> 24:55.680]  That's hard to do and it requires space.
[24:55.860 --> 24:59.820]  At the very least, it requires capacitor space for a switching regulator.
[25:00.520 --> 25:07.120]  We tried doing this in our prototype and it ended up being huge, as you'll see in a second.
[25:09.820 --> 25:11.920]  Next, let's talk communication methods.
[25:12.600 --> 25:16.140]  We're going to be talking to a different device here.
[25:16.360 --> 25:21.520]  A reader device, as this was in our gold sheet previously.
[25:23.580 --> 25:26.820]  Again, we can't just have a screen on the side of our arms 24-7.
[25:28.180 --> 25:31.580]  Even with communication methods, power is very important.
[25:31.860 --> 25:42.240]  In fact, power will often use more battery or more wattage than just taking measurements.
[25:42.760 --> 25:44.520]  Especially for things like Bluetooth.
[25:45.020 --> 25:52.320]  If you don't use Bluetooth low energy, you're going to be absolutely draining the power of your battery very quickly.
[25:52.320 --> 25:58.360]  And it means you're either going to be taking measurements really infrequently because you're uploading the data all the time.
[25:58.720 --> 26:04.200]  Or it means that your battery life is going to be really short and needs to be replaced all the time.
[26:05.460 --> 26:09.860]  Some of these are just general goals for any device, but they especially apply for a medical device.
[26:10.280 --> 26:12.160]  They should meet regulatory standards.
[26:12.260 --> 26:16.660]  You want the FCC and the FDA to actually approve it so you can sell it.
[26:16.660 --> 26:19.740]  So it should be in the industrial scientific medical band.
[26:20.400 --> 26:26.780]  Literally everything that someone would normally buy will be in the ISM band, which is just...
[26:26.780 --> 26:30.220]  The standard one would be 2.4 gigahertz, same as Wi-Fi.
[26:31.220 --> 26:33.200]  Next, the protocol needs to have strong encryption.
[26:34.060 --> 26:37.100]  Men and middle attacks and glucose monitors do not mix.
[26:37.580 --> 26:38.540]  Not very good.
[26:38.780 --> 26:40.200]  Especially for the manufacturer.
[26:41.860 --> 26:45.000]  And it needs to be cheap and easily manufacturable.
[26:45.000 --> 26:53.600]  So something like Bluetooth, despite its higher power consumption, Bluetooth chips are really easy to come by and can be made really cheap.
[26:53.780 --> 26:56.080]  That's how knockoff AliExpress AirPods are.
[26:56.740 --> 26:58.740]  Because they can make a huge profit on them.
[27:00.220 --> 27:03.340]  So this leaves us with only a couple options remaining.
[27:03.340 --> 27:06.240]  The three that I picked out were Bluetooth, LoRa, and Zigbee.
[27:06.460 --> 27:10.740]  Zigbee would be a little weird of a choice because it's mainly made for IoT.
[27:11.100 --> 27:12.840]  But it would theoretically work.
[27:13.600 --> 27:17.280]  We personally use LoRa in our design.
[27:19.640 --> 27:24.840]  Next, if any of you have done any hardware projects, we know about application notes.
[27:25.220 --> 27:28.060]  They're super, super, super useful.
[27:28.960 --> 27:30.180]  There's different op...
[27:30.180 --> 27:40.480]  There's essentially a bunch of different manufacturers for chips, and they'll all publish data on essentially what you want to make.
[27:40.900 --> 27:44.200]  Here I have examples on what we're doing, which is building a glucose monitor.
[27:44.560 --> 27:47.880]  They serve as a good baseline for what you want to do.
[27:50.220 --> 27:57.460]  So even developing software for it, picking out the right chips, figuring out how it works even on a chemical level.
[27:58.500 --> 28:07.020]  So a good baseline just for like finding the right chips, finding the right documentation, is to look at...
[28:10.120 --> 28:11.680]  We can develop a block diagram.
[28:11.940 --> 28:15.960]  I made this super quickly for essentially like a bare bones glucose monitor.
[28:15.960 --> 28:21.100]  All it's going to do is take in a sensor reading and send it out to an antenna.
[28:21.540 --> 28:24.460]  So we have our sensor array connected to a microcontroller.
[28:25.320 --> 28:28.480]  Some microcontrollers will have integrated communications.
[28:28.540 --> 28:29.600]  Others won't.
[28:30.580 --> 28:42.640]  So in our specific design, we found that a lot of low power microcontrollers don't because they essentially don't expect you to use wireless communication.
[28:42.640 --> 28:44.280]  It uses up too much power.
[28:44.600 --> 28:49.260]  So we have a separate low power communications chip connected to an antenna.
[28:49.740 --> 29:07.280]  And to power it, we either have our battery or our new experimental power supply, being like movement or heat, which will then go into, say, a battery management IC, or a step-up regulator.
[29:09.960 --> 29:12.560]  Then it was time to actually make the thing.
[29:13.000 --> 29:14.120]  That took a while.
[29:15.580 --> 29:21.660]  Step number one in that, forget that you have any, even any shrivel of understanding in electrical engineering.
[29:22.060 --> 29:29.940]  Again, not to make excuses, but I don't exactly have that much knowledge in the field of designing circuit boards.
[29:30.200 --> 29:32.020]  We wanted to make something anyway, though.
[29:32.260 --> 29:33.200]  So we did it.
[29:34.160 --> 29:38.680]  Next, you want to fire up your PCB software of choice, which those things can get really expensive.
[29:39.680 --> 29:40.460]  KeyCat is great.
[29:40.700 --> 29:41.300]  It's open-source.
[29:41.500 --> 29:42.380]  Highly recommend it.
[29:42.660 --> 29:50.460]  But there's other alternatives, like Autodesk Eagle, or if you actually have a job as an electrical engineer, Altium Designer.
[29:51.680 --> 29:54.460]  After that, you want to read up on a ton of TI and ONCE.
[29:54.660 --> 29:55.620]  A ton of TI.
[29:55.920 --> 29:57.220]  We use TI and ONCEMI.
[29:58.240 --> 29:59.360]  That was our example.
[30:00.400 --> 30:07.740]  But outside of that, find a chip manufacturer, read up on all the documentation you can, and try to turn it into your own schematic.
[30:08.700 --> 30:10.300]  That takes a really long time.
[30:10.600 --> 30:15.380]  It took us essentially, like, literally until the very end.
[30:18.320 --> 30:20.220]  After that, you get to make it.
[30:20.500 --> 30:22.180]  This is our chem lab.
[30:22.280 --> 30:24.760]  So this is where we were doing all this stuff with the glucose oxidase.
[30:26.240 --> 30:28.780]  After you make a physical...
[30:28.780 --> 30:34.200]  Sorry, after you make a digital file of the thing, you have to turn it into a physical product.
[30:36.000 --> 30:38.560]  First of all, that involves finding a PCB manufacturer.
[30:39.120 --> 30:43.580]  Again, thanks to China, you can get PCBs for literally like 50 cents a pop.
[30:43.580 --> 30:44.400]  It's great.
[30:45.640 --> 30:48.200]  We looked at a website called JLCPCB.
[30:48.420 --> 30:49.760]  Again, highly recommend them.
[30:50.900 --> 30:55.720]  And we put in our file and sent it off to them.
[30:55.800 --> 31:00.520]  And about 10 minutes later, they told us that the via size was too small.
[31:01.140 --> 31:07.860]  A via, for those who don't know, is just a little hole they drill into a PCB to connect multiple layers.
[31:07.860 --> 31:15.300]  So circuit boards will have sometimes even four, eight, 16 layers that will need to be connected together.
[31:15.500 --> 31:24.620]  And so they'll drill us very small hole, literally like one, two millimeters, fill it with copper, and then it'll connect the layers.
[31:26.840 --> 31:39.000]  After fixing all of those problems, which took a lot of time in of itself, we had to pay around, it was like $4 for the PCBs themselves.
[31:40.580 --> 31:44.180]  Again, the price of stuff in China is just mind boggling to me.
[31:44.440 --> 31:53.960]  But then to not get it eight months from now, especially after learning the mistake of the glucose oxidase, we spent like $70 in shipping, so it doesn't take forever to arrive.
[31:53.960 --> 32:03.260]  So we ended up paying like just way more in shipping than even the price of this goods itself, but it was still worth it.
[32:04.260 --> 32:08.060]  After that, boards arrived, and we started soldering.
[32:08.300 --> 32:16.940]  We had just a giant line of parts that we ordered, and we'd assemble them piece by piece.
[32:16.940 --> 32:24.100]  It took three hours or so just to do like two prototype, two non-functional prototype boards.
[32:24.540 --> 32:31.060]  And let me tell you, Mr. Kane, he found it really cool, but I think he wanted to go home by then.
[32:35.080 --> 32:38.660]  So, like I was saying before, this is a non-functional prototype.
[32:39.300 --> 32:44.620]  I do not have any, you know, undergraduate or graduate experience in electrical engineering.
[32:44.620 --> 32:47.900]  I don't... or even really that much hobbyist experience.
[32:48.120 --> 32:50.120]  But I wanted to try to get a start.
[32:50.840 --> 32:54.080]  So PCB software generally, really complicated to use.
[32:55.060 --> 32:57.300]  I used Google for the tolerances.
[32:57.300 --> 33:05.650]  I feel like a real EE might, you know, actually try to figure out what's good for their PCB manufacturer.
[33:06.520 --> 33:07.480]  It really didn't.
[33:07.860 --> 33:14.200]  And also, soldering like a madman, like I said in the last slide, means that there's a million short circuits on the board.
[33:14.200 --> 33:18.520]  So, if we plug that thing in, I don't really want to see what happens.
[33:20.320 --> 33:23.600]  After that, fresh PCB out of the oven.
[33:23.880 --> 33:28.660]  We have two ones that were made.
[33:28.660 --> 33:31.360]  I actually have them here if you want to see them.
[33:32.060 --> 33:33.120]  Pass them around.
[33:34.410 --> 33:35.700]  These are huge, by the way.
[33:36.180 --> 33:40.560]  And the reason why they're so big is because we need to power them.
[33:40.700 --> 33:43.660]  We power them with, like, thermal electric energy.
[33:43.660 --> 33:44.820]  So you have...
[33:46.320 --> 33:48.080]  This would be the real close monitor.
[33:48.260 --> 33:51.060]  It would be attached to you.
[33:51.420 --> 33:59.680]  And the temperature difference between your body and the ambient air would produce a very small amount of current, which we then use to power it.
[33:59.740 --> 34:01.460]  That's why we have all these huge capacitors.
[34:01.740 --> 34:04.900]  So you guys can see how horrible of a job I did soldering everything.
[34:06.500 --> 34:07.900]  And this was...
[34:07.900 --> 34:12.900]  The smaller board is supposed to be almost like a receiver board.
[34:13.040 --> 34:18.220]  You'd plug it into USB, and you'd be able to download the data from another device.
[34:23.860 --> 34:26.660]  That's a consequence of having used Lara instead of Galileo?
[34:28.440 --> 34:29.800]  No, not actually.
[34:30.000 --> 34:36.320]  What we wanted to do, though, is just have two different boards to essentially show for the competition.
[34:36.380 --> 34:37.220]  It was fun.
[34:37.380 --> 34:38.420]  I wanted to do it again.
[34:39.140 --> 34:39.920]  We had...
[34:40.980 --> 34:43.640]  I had time to design it, so...
[34:43.640 --> 34:44.320]  I thought I might as well.
[34:46.220 --> 34:47.520]  And that's really it.
[34:47.520 --> 34:49.140]  I know I spoke for a while.
[34:56.740 --> 34:57.480]  Any questions?
[34:59.380 --> 35:03.620]  So you don't want to go into manufacturing and money units?
[35:04.040 --> 35:08.760]  Well, for that I'd have to get FDA approval, and I don't think the FDA would let me do that.
[35:09.580 --> 35:10.580]  To say the least.
[35:22.260 --> 35:23.480]  All right, then.
[35:23.880 --> 35:27.660]  Maybe we can have a DIY glucose monitor get approved.
[35:27.940 --> 35:28.960]  Can you repeat it?
[35:29.440 --> 35:37.320]  Oh, he was just saying that the FDA, if you're not profiting, will give you what, like a fast track in a sense?
[35:45.580 --> 35:46.040]  Yeah.
[35:46.260 --> 35:48.680]  And they'll help you if your product isn't for profit.
[35:52.060 --> 35:57.160]  The question is perhaps slightly outside the scope of what you did about the design of continuous monitors.
[35:58.180 --> 36:06.820]  Presumably the enzyme, the glucose oxidase, and the hydrogen peroxide aren't things that are hanging around in your bloodstream.
[36:06.820 --> 36:09.120]  No, they're not.
[36:10.920 --> 36:18.560]  I think what some manufacturers have done is essentially have like a vinyl coating that will separate it from the rest.
[36:18.680 --> 36:21.340]  So you're not going to be getting hydrogen peroxide actually in your bloodstream.
[36:21.340 --> 36:31.080]  It's more taking a sample of blood and then in almost an enclosed environment turning that into hydrogen peroxide.
[36:32.380 --> 36:35.040]  That was what the foam was used for in our example.
[36:35.220 --> 36:38.040]  Obviously that wouldn't work in real life, but...
[36:38.040 --> 36:45.060]  I don't know if this would have worked, but animal blood or other things like that, did you think about testing it on anything else, like real?
[36:45.420 --> 36:46.400]  Well, I don't know.
[36:46.500 --> 36:48.660]  I think our school might have some problems about that.
[36:51.460 --> 36:54.140]  Is the enzyme consumed making the hydrogen peroxide?
[36:54.400 --> 36:54.680]  Yes.
[36:55.960 --> 36:58.880]  That, I think, is actually one of the reasons for replacement.
[36:59.120 --> 37:12.980]  So even if you were to have, say, a movement-powered or heat-powered device that wouldn't have a battery, so you wouldn't have to replace the battery for a long time, what you would have to do is replace it because of the glucose oxidase.
[37:16.780 --> 37:33.020]  And the... I've never seen, or you, the on-arm, the continuous glucose monitoring, but... so it actually sits inside a stream of... I want to say a vein or an artery of it.
[37:33.140 --> 37:33.160]  Yes.
[37:33.320 --> 37:36.100]  I mean, or is it just in between, like, regular system cells?
[37:36.100 --> 37:36.780]  No, it's a shot.
[37:37.540 --> 37:37.940]  Right.
[37:38.560 --> 37:41.600]  It's actually there's a interstate with fluid, and that's actually with glitter.
[37:41.820 --> 37:44.220]  It's a bit like some of the items behind the...
[37:45.080 --> 37:45.480]  Okay.
[37:45.920 --> 37:47.240]  Is there microneedles on those?
[37:47.380 --> 37:47.880]  There are.
[37:48.600 --> 37:48.940]  Hmm.
[37:50.620 --> 37:51.540]  Anyone else?
[37:53.700 --> 37:57.000]  How much did it cost over, like, once you assembled everything?
[37:58.140 --> 38:03.620]  Well, the glucose oxidase was around $90, but we got around two kilograms of it.
[38:04.300 --> 38:07.920]  We... you can't really buy a tiny amount of glucose oxidase.
[38:09.020 --> 38:14.960]  So, Mr. Kane has 1.9999 kilograms in his room now.
[38:15.280 --> 38:17.920]  It's a... it's a gift from the SkillsUSA team.
[38:18.060 --> 38:19.240]  We'll see what he uses it for.
[38:20.260 --> 38:21.500]  So, that was $90.
[38:22.480 --> 38:24.700]  With shipping, the...
[38:26.140 --> 38:28.860]  the PCBs were around $70 or $80.
[38:30.860 --> 38:35.480]  And I think parts... we bought, like, a ton of extra parts.
[38:35.480 --> 38:39.760]  And that was around $120, $150.
[38:40.340 --> 38:43.320]  I think our overall budget for the project was $300.
[38:43.500 --> 38:45.820]  And we came pretty close to that.
[38:45.940 --> 38:49.360]  But that includes some other things, like posters and whatnot.
[38:49.720 --> 38:50.020]  So...
[38:52.480 --> 38:54.440]  A lot better than $15 a day.
[38:58.700 --> 39:10.700]  Considering the challenges in miniaturization of batteries and power, have you considered using something like, let's say, induction or NFC through a separate grid power source?
[39:10.900 --> 39:17.660]  So, basically you have the sensor itself, but then it has a core oversteering through a power source.
[39:17.800 --> 39:19.620]  That's also, let's say, belt mounted.
[39:20.060 --> 39:21.860]  We actually discussed that.
[39:22.840 --> 39:30.760]  What was interesting about that is it would last for a much longer time than, say, even like a movement or a heat powered one.
[39:31.140 --> 39:36.580]  Because, you know, those devices don't work outside of optimal conditions.
[39:36.580 --> 39:38.980]  If you're sleeping, you're not moving as much.
[39:38.980 --> 39:40.320]  You're not generating as much power.
[39:40.820 --> 39:50.360]  With NFC, you kind of have a similar problem, but it'd be less exacerbated where you have to power it up every so often.
[39:50.360 --> 39:55.620]  But I guess what you could do is have almost like a rechargeable battery.
[39:57.100 --> 39:59.180]  NFC, also for like...
[39:59.180 --> 40:03.440]  Are you thinking like, say, a phone to this or like a separate brick?
[40:04.880 --> 40:11.480]  Yeah, I feel like that would work, but you'd have to have some sort of separate power source.
[40:11.720 --> 40:23.320]  And at that point, it might be more of an inconvenience to have to hold a brick against you like you're charging a phone than to, you know, just take a glucose... put a new glucose monitor.
[40:23.640 --> 40:24.560]  I don't know.
[40:24.760 --> 40:26.340]  I haven't done product tests.
[40:26.640 --> 40:29.580]  But overall, yes, definitely.
[40:29.580 --> 40:31.460]  I think that would work and it would be a good idea.
[40:36.740 --> 40:37.720]  I'm just curious.
[40:39.180 --> 40:43.880]  Do you see this applying to whatever you want to do as a career?
[40:45.660 --> 40:48.540]  Yeah, I really feel like I want to go into engineering.
[40:49.720 --> 40:51.500]  So, I'm a senior right now.
[40:51.700 --> 40:52.940]  Next year, I'll be going to college.
[40:53.220 --> 40:56.400]  Hopefully, I can do something along the lines of electrical or computer engineering.
[40:56.760 --> 41:07.140]  I mean, this was really like almost a test in how much do I like engineering, how much do I like engineering design, and the specifics of electrical and computer engineering.
[41:07.360 --> 41:13.360]  One of the original goals of this was also to develop some form of software for it, but we just didn't have the time.
[41:14.260 --> 41:15.780]  But, yeah, definitely.
[41:16.040 --> 41:19.080]  I'd love to go into some sort of engineering field.
[41:22.660 --> 41:23.820]  Yeah, yeah.
[41:28.050 --> 41:36.190]  One of the interesting things about those new glucose monitors, I've actually seen people using those like off, almost like off medical thing.
[41:36.670 --> 41:39.410]  They're using for like cheating and like athletics.
[41:40.290 --> 41:44.030]  They're using it on athletes to figure out their glucose when they're running and stuff.
[41:44.450 --> 41:45.750]  Like bike racers.
[41:45.750 --> 41:47.030]  It's kind of unusual.
[41:47.850 --> 41:51.170]  Hopefully, those people have got cheap stuff off prescription.
[41:51.590 --> 41:51.930]  Yeah.
[41:51.930 --> 41:53.850]  They actually use them for like training and stuff.
[41:56.550 --> 41:57.850]  Yeah, hopefully they do.
[41:57.990 --> 41:58.590]  Hopefully they do.
[41:58.770 --> 42:03.030]  I feel like as technology improves, we might see slight price reduction.
[42:03.350 --> 42:10.710]  But overall, we're not going to be seeing the sort of major changes.
[42:10.710 --> 42:15.150]  Because really there's only a few glucose CGM manufacturers.
[42:15.550 --> 42:19.490]  They're not really like in a giant price war.
[42:19.990 --> 42:21.210]  They're pretty stacked.
[42:21.430 --> 42:25.310]  We're not going to be seeing major changes in the prices of these things.
[42:25.370 --> 42:27.170]  But they will slowly go down.
[42:29.930 --> 42:36.050]  Did you by any chance come across anything around like non-invasive continuous monitoring?
[42:38.030 --> 42:39.090]  I've seen it.
[42:39.210 --> 42:41.090]  That wasn't really the focus of our research.
[42:41.250 --> 42:44.630]  It was harder to find any research about that, frankly.
[42:44.910 --> 42:47.310]  But it would be something that I'd like to research further.
[42:47.610 --> 42:49.350]  Yeah, I think it's still pretty cutting edge.
[42:49.650 --> 42:49.990]  Yeah.
[42:53.070 --> 42:54.650]  Well, I know I'm a little bit early.
[42:54.850 --> 42:56.810]  But thank you all for attention.
[42:57.030 --> 42:57.050]  Oh.
[42:57.050 --> 42:57.550]  So...
[42:57.550 --> 42:59.300]  A gentleman there was mentioning...
[43:00.590 --> 43:05.290]  Have you considered collaborating with the biohack community?
[43:05.750 --> 43:09.850]  Because that is something that is definitely within their scope in terms of...
[43:09.850 --> 43:09.910]  Yeah.
[43:10.210 --> 43:13.550]  ...measuring their own internal performance data with respect to their body.
[43:14.150 --> 43:20.290]  So potentially they might have a much larger troll of knowledge and insight that might be able to help you around your project.
[43:20.530 --> 43:20.830]  Yeah.
[43:20.930 --> 43:22.200]  So just going back to that...
[43:25.480 --> 43:26.780]  ...and starting this.
[43:26.780 --> 43:36.580]  So I feel like the project isn't that far enough along to really get outside involvement yet.
[43:36.640 --> 43:53.480]  But overall, it would definitely be a good idea to just get help and especially get the view of people that are diabetic or have a continuous glucose monitor to just hear what they have to say about their daily use of it.
[43:58.390 --> 43:59.070]  Yeah.
[43:59.250 --> 43:59.830]  Yeah.
[44:00.810 --> 44:01.290]  Give
[44:05.060 --> 44:06.160]  us a bit of a...
[44:06.160 --> 44:06.400]  Sure.
[44:06.780 --> 44:08.520]  ...run-down of what's in the board.
[44:09.420 --> 44:10.960]  Couldn't probably use this so I didn't.
[44:13.840 --> 44:14.320]  Well...
[44:14.320 --> 44:15.180]  Okay.
[44:16.800 --> 44:17.280]  So...
[44:17.280 --> 44:18.680]  The board...
[44:18.680 --> 44:20.260]  This is just the carrier...
[44:20.260 --> 44:20.500]  Sorry.
[44:20.580 --> 44:23.720]  This is just the companion board that will hook up to a computer.
[44:24.320 --> 44:28.060]  All it really is, is a JTAG port for debugging.
[44:28.380 --> 44:30.820]  Again, we were originally going to flash some software onto it.
[44:30.820 --> 44:32.340]  But we just didn't have enough time.
[44:33.860 --> 44:42.880]  And also, just a general microcontroller, communications chip, and a really small chip antenna from Coilcraft.
[44:43.180 --> 44:47.200]  And then just a bunch of passives for the antenna, passives for USB power.
[44:47.200 --> 44:49.960]  So it's a really simple device.
[44:50.180 --> 44:54.200]  It really doesn't have anything outside of...
[44:54.760 --> 45:01.420]  USB goes in, microcontroller processes, antenna data goes out.
[45:01.420 --> 45:15.020]  The bigger device, the bigger square monitor, that has a bit more going on because it has thermal power.
[45:15.300 --> 45:17.840]  That's why it has the huge capacitors.
[45:18.080 --> 45:27.340]  And there's also a couple more chips specifically designed to help regulate the power and to help boost it up to the correct voltage.
[45:27.340 --> 45:35.000]  Whereas with USB, you know, you're going to be getting noise, but it's not going to be nearly the same problems.
[45:41.400 --> 45:42.340]  Any other questions?
[45:45.160 --> 45:45.740]  All right.
[45:46.540 --> 45:48.140]  I know I'm a little early, guys.
[45:48.560 --> 45:50.500]  But thank you all for attending.
[45:50.760 --> 45:51.820]  I hope you had a good time.