 Thanks for being here that early in the morning. I'm going to talk about a motion sensor, a wireless motion sensor. My name is Tere Gronet. I'm from Tenzibal Display. I should speak louder. Okay. I'll try my best. So, in my company, Tenzibal Display, we build giant multi-touch screens. That's how we make money. We sell them to museums. We sell them to luxury brands and things like that. But we also all have a musical background. And we like to play with objects, Tenzibal things. So, we made this one that's probably going to remind you of the red table. This one is an open source version. It's called Tenzibal. And basically, you can control music with objects that are detected by the table. All of these objects, they have a unique identifier. They are recognized by three little bits that are simulating three little fingers. Each object is unique and we know the orientation and we know the position. But when you take out the object from the table, from the two-dimensional surface, you cannot interact with it anymore. So, what we thought about was adding a little sensor. So, we could interact even outside of this two-dimensional surface. So, to explain a little more, the control different track. Each track can be modified or modulated thanks to the parameters. But adding this sensor, we had this problem, we looked at what exists and we found that everything was either way too expensive, way too consuming, or too complicated to use. So, we just gave up trying to find something and made our own. So, normally I would do a demo right now, but I've been a bit shortened on the tank. So, I'll show you a little video after. And if you want to see me, I can also show you with my phone. So, basically, I went to this festival called Burning Man. I've been quite amazed by all of this creativity there. And one of the projects was a huge ball of flame. And my little dream is to control flame with motion. So, this object allows you to control the motion. Any media out of the motion. So, in this case, controlling with a sensor, each of the motion could trigger a flame, or light, or music, or effects. Just an example that I want to try. But there's a lot of other possibilities. So, some people came to see me and talked about serious things like near-realization, artistic analysis from bridges, like civil engineering, and a few other applications that put it on your drone and see what happens, put it on your skateboard. And from this data, you can control things like the light of the room, the sound of the volume of the music that you're listening to. Can you imagine putting it in maybe a candle and just control the volume of the music or the light? Plenty of other things. A very serious application, control the game. So, that's a way of using it. You can see that orienting the sensor is sending another fake mouse. That's basically the idea. So, how did we do that? As I said, we started by testing different development boards. So, the usual suspects are the famous Tinti, kind of Arduino company, but smaller, cheaper and quicker. A little motion sensor board and a wireless module. The three boards, if you put them in one, you get one main twist. Same scale, one inch square. It's the size of a coin cell battery. And basically, I showed you the two sensors on the board that was on the left, and the wireless module and the processor in this chip. So, to server, we have this new feature, which is Bluetooth through energy that allows communicating with smart phone, tablets and computers also. We have lower power, really small size, autonomous batteries at the bottom. So, you can be independent with that. It's slower than its open source. If anyone is interested, I'm happy to share with you some details to collaborate. A little quick explanation of what's happening inside of the PMR in the processor that is embedded on the board. We have two kind of, three kinds of sensor. Accelerometer, everyone knows. There's a magnitude meter and the gyroscope. These are doing a bit of the same thing, but not exactly. I think that's the most important part. The magnitude meter, he knows roughly where the north is, but it's not very accurate. It's quite slow. The gyroscope knows where you're moving, but when you're not, it doesn't know energy. So here, we do a 90 degrees rotation from, let's say, north, and then east and north again. And that's the same. So you see only when it's moving on the gyroscope, and when it's moving back, to get the position from the velocity, the angular velocity, you integrate the data, which is an accumulation in the discrete work. So you do this integration, or accumulation, which adds all the noise. This noise adds what we call a drift. So basically, when you start, you are at zero degree, 90, and when you come back to zero, the drift makes that you don't know where to go. So you kind of fix that. We use the magnitude meter by merging them smartly and get something nice. So quickly, we make the fusion on board, send it over to the energy, receive it in an application, and send it to an application like Ableton Live, the musical controller I was showing you. Quickly, I made this hardware using a printer. It's a kind of Google Doc for PCV. I send it to the factory. It's got 20 processors, fake, or 40. I went to my hacker space. Someone was working in a hospital. In the hospital, you made an x-ray. We looked at all the details. Apparently, the problem was in the processor. So some more hacker space pictures just to explain if you ever want to do this kind of surgery. You can just warm it up with a heat gun to remove pain, or you can use better tools like the real official heat gun, or you can just cruise them with a normal hotbed. Actually, it works pretty well. I could show you a video if you want afterwards, but timing is short. Quickly, to avoid that kind of problem, testing. That's the conclusion. Sorry, if you want to speak with Cedric, please. So, excuse me. You'll be here the rest of the day. Yeah, we'll be here. Okay, so please find a minute. Thanks very much.