 Thanks for having me here. I just literally came from teaching just a few minutes ago. So a little bit about me, I'm an innovation manager at Intel. And what that means is I take Intel products and I put them in places they were never meant to go and see what happens. And a lot of times, really interesting things come out of that. And this is the story of one of those interesting things that really made me think of transhumanism and cyborgs and whatnot. It really is about scoliosis and the future of 3D printing. I had a little bit of an epiphany around this when I basically crossed internet of things and 3D printing and machine learning in all in one space. So that's what I'm going to talk about today. First, a little bit about 3D printing. How many here are familiar with it? Yeah, almost everyone is. The most common type of 3D printing is where you deposit layers of molten plastic. You use a gantry system. And the device builds an object and layers. There's other types of 3D printing as well, though. And a lot of times what people think of in 3D printing are these little trinkets, like the owl on the log and just toys and castles and all that kind of fun stuff. But there's really interesting work happening in finished 3D prints, finished goods, such as this. How many of you have seen this? This is a titanium vertebrae printed by a powder bed printer. And this is an implant that actually went into a person. And thinking about 3D printing as a means to get to a final product, these kinds of applications are incredible, but they're one-offs. And you've got to wonder, how do we apply what we've learned here more broadly? Well, Amber had talked a little bit about my work with the Enable group. This is the group that 3D prints prosthetic hands for anyone in need. And this uses that layer of molten plastic. And it's an amazing application of this technology. And imagine every one of these hands is custom made to the user. And we don't think anything of it. To customize the scale, the fit, the form, to whatever the user needs, it's a sign of what's to come. And it's just one of the most amazing things I've been part of. If you have a 3D printer, consider joining the Enable group and making a hand for someone. It's a self-organized group of 10,000 people that are now making those hands for anyone who needs them. But I'm not going to talk about that so much. I'm going to talk first about scoliosis. How many here know what scoliosis is? How many know someone who's been affected by it? Amazing. Every time I do this, it's about a third of the room. Here it's much more. It's an abnormal curvature of the spine. And it usually affects girls 12 to 14 years of age. And what happens is at the onset of scoliosis, the physicians and the girls have a choice. They can wear a brace or they can face surgery. And even at the end of wearing a brace, they may have to face a surgery. So this bracing technology is actually pretty old. It's been around for hundreds of years. It started as the Stillman brace was basically all leather. There's the Milwaukee brace, which was created in the early 20th century. That's made out of thermoplastics and metals. And this thing is really heavy deal. You have to wear it on the outside of your clothing. And then the Boston brace came along as the first brace that actually was based on data analysis. And this is also thermoplastic, but it's much more conformal to the user's body. If you notice, the form hasn't changed, but the materials have over time. So as an innovation manager, I got to meet a company called Unique that is making a 3D printed scoliosis brace. So again, the form is exactly the same. But they took a 3D body scan and then used 3D printing tools to make this amazing conformal brace. But you can see some of the things that are different. Like for example, there is a design pattern in it. And there are things like hinges inside of there. And what that allows us to do is make something that's more acceptable to the user. And we get the complexity of these designs for free because of 3D printing. So what they've really done with this brace, more than anything else, is they've changed the discussion away from stigma to style. This brace is something the girls want to wear. But there's still a problem. The brace is covering up scoliosis, or it's attempting to remediate scoliosis, and then needs to be worn 18 hours a day. How many of you have a 12 to 14-year-old daughter? OK, how many of you have teenage daughters? OK, you can't make them do anything for 18 hours a day, 18 minutes a day, and maybe. But so how do they solve this problem? The answer really comes down to surveillance. So when I found unique, they were putting this data logger on the scoliosis brace, and it measures temperature, the temperature of the body. So you can make a rough assumption that if the sensor is seeing heat, that it's being worn. Guess what happens? 12 to 14-year-old girls are really, really clever. They take this brace off, and they put it in the sun. They put the dog in the brace. They do all these wonderful little things. So being Intel, we were like, we probably have a microprocessor for that. And sure enough, we did, because we called it Curie. And it was this neat little moat that had Bluetooth integrated, sensors integrated. And it had charging points. It was a great little device. And what we did is said, we are going to make these girls wear this brace for 18 hours because they're going to know which 18 hours in the day that they're going to wear it. And we was all wrong. Our assumptions were so bad, it was just ludicrous. So we went and actually talked to the mothers and daughters, and they figured out that they just couldn't agree on which 18 hours a day to wear it. So our sensor had Bluetooth in it. And we made an app that you see over here. And this app records exactly how long during the day it was worn, over a month and over, I think it was a year in there. I can't see this close. But the end result was that the argument stopped. And the brace was being worn in a conformant way. It was great. But that's not really why I'm here to talk. I mean, we had all sorts of sensors in there. We got the outcome we were looking for. But I want to point you to those three dots. Do you see those there? Those are three pressure points on the brace. And that represents, the green represents that the pressure is just right at that point. Kind of like a baby seat, when you put a baby in, you don't want to cinch it down too tight, because you might harm baby. The first time that the brace is put onto the user is the only time that it's tight enough. And now we were giving indications that it was tight enough for you. What happened next was magical. The designers of the brace said, can we have that data? We want to use that thing. We want to use that for checking our model. And what they did with it was this. So this is apparently not going to run unless I click. And maybe not even then. Studio Bitanti is an industrial designer in New York City that we paired up with Unique. And they took the data of the brace wearer and they removed the plastic that wasn't therapeutic. They did an optimization. First, a first pass of a finite element analysis, then they used the data to make the brace into a new form. And what they left with was a wearable work of art. Imagine that this thing has the same efficacy as the one you saw before. And it changed the discussion yet again. So this is a brace that not only is functional, but the girls actually really want to wear. And I had a bunch of other pictures in here. We demonstrated this brace at a fashion show at the White House. It's been on the runway fashion show at Obama's White House because you know what will happen if we tried this here. You get the idea. It really became a symbol of empowerment. But we got to thinking as Intel, how do we sell a little bit more microprocessor? And it became a symbol of mass customization. And that's what we're going to talk about here for the rest of this. There's a lot of friction in the design process of making a scoliosis brace. Three doctors visits, countless hours of design time and of molding time, and then production on top of that. There's really seven manufacturing steps that we could remove. So we got to thinking, OK, we can make a better 3D printer. We can make better sensors. Great. We can make that happen. I'm not going to talk about that here. What I want to do is think about the design process holistically. Right now it's very linear. You have embedded sensors. You take sensor data into a design tool and you do 3D printing. That's where we want to go. Today we don't even have the, or before we did this, we didn't have even the sensor part. So we got to thinking, this is what a designer has to deal with when they're looking at sensor data. It's waveforms. What is the user trying to do here? Without me putting tags on here, it would be really difficult to see that the user is trying to lie down and get comfortable. The bottom row here is the orientation, up, down. And then those are the three pressure sensors where the pressure is spiking, the scoliosis brace is digging into the user. So we got to thinking, well, what if we were to take this data and feed it into a machine learning system and let it learn what the user is trying to do? So today, you feed in sensor data, you update an app screen, you may personalize what a user sees. Tomorrow, those should translate into design hints. And this would be data coming straight off of the scoliosis brace. But the magic happens where you take all the data that's surrounding us. Our phones are vast repositories of what we're trying to do, of our intent and of our lifestyle. And what if we were to mix this in and turn this vast repository of data into the ability to customize for lifestyle? Effectively, we'd be closing the loop on design and making it easier for us to get to solving problems and customizing at scale. So think again of the waveforms. If we married that to phone data, and we realized that we're only seeing these waveforms at night when we pair that up to our Apple sleep-sensing platform, and we realized that the user is trying to go to sleep and the brace is cutting into her. They have to wear this while they're sleeping, by the way. I don't know if I mentioned that. We can do something about that. We can optimize for their lifestyle by changing not only the function of the brace, but the form of the brace, the materiality of the brace, the architecture, just using the sensor data. And what's insane about this is it's happening. There is an API now where you can upload a body scan and get a Scoliosis brace model out. That's how unique is developing their Scoliosis braces today. It's through the Francis Botanti Genesis API. It's insanely cool. So that's where I come to my big point, is that we can get to hyper-personalization, customization for lifestyle by marrying IoT and machine learning and 3D printing. So that's my talk. I have one more little bit of advice to give. If you want to talk about this stuff, I'll be available. And you can get at me on LinkedIn or on Twitter. Now, if you want to try your hand at Generative Design, which is what this comes down to, there's a tool that you can use today. It's called Bots 101. If you go to their app, you can write in relatively conversational English a description of a model and it will generate a 3D printable file for you. It's incredible stuff. So thank you for having me and get some open to questions.