 Thank you very much for attending this new conference. Good afternoon. I'm going to start and explain to you how my company, ST Microelectronics, can help you to bring innovation in home. First of all, do you know we are on an embedded world now. We talk about 32 bits, we talk about digital, but we are all analog, you and me. Our sensing is ear, taste, smell, touch, feel, see. So what is crucial to bring innovation is to bring innovation with sensors. It's one of my topics today. You see that sensors are old, okay, exist for a long time already. And each new progress in the sensing, they bring new applications, new technologies, new markets. Talk, listen, gesture recognition, speech recognition is one of the topics I will tell you today. What is important as well for us in ST is to help the startups, the small companies to go to the market. They have an idea, they want to go to the market very fast, and we want to help them to innovate, to go very fast to this market. For device prototyping, for the software, we bring all the bricks, okay, building blocks. And then we give the technical support to go there. I don't know if you heard about the Nucleo, Nucleo boards. It's a very small and cheap MAMC boards. We also provide expansion boards for sensors, for RF, for communication, motor drivers, NFC. And then you build your blocks, hardware blocks, and we provide also some software. This is some details about the hardware blocks we have today. And we will have much more this year, we will release a lot. Okay, STM32 Open Development Environment is not only hardware, it's also software. But it's concerning hardware, it's Nucleo little boards, white boards, as you see, the white one here. Then you have Bluetooth low energy here, sensors, but also available NFC, motor control, and flight sense for proximity sensing. In terms of software, we provide of course the very low, lower layer software, abstraction layer. This is provided by our tool, which is called STM32 Cube, on which you can generate the initialization of the microcontroller. Then even for application, we bring some additional bricks like board support package, middleware libraries like sensor fusion, which is a very high difficulty, high difficult board to make, software to make. We bring it to you free of charge, and also some examples of software. So if you want to build an electronic compass, then we bring to you the software example, the sensor fusion, the accelerometer, gyroscope, magnetometer together, data compensating the drift of the sensor, and also giving you all the data necessary for your application. What is interesting is also to go further in home automation, okay? It doesn't work anymore, sorry. Okay, home automation. So for each application of home automation, we have the sensors, we have the processors, we have the power management, and we have the RF link. So you can have different applications, but we always have the bricks and the software. Okay, so these are the products that we have in ST to process, to sense, to connect, and to power the applications with some examples. So in the field of connectivity, we propose Wi-Fi module, Bluetooth modules, Bluetooth low energy ICs, and sub gigahertz ICs, and also NFC. For power management, we propose energy harvesting, and we have a demo here on the booth of a battery-less pedometer. So it's a pedometer that works without a battery. Just the shock like this of the shoe on the floor gives enough energy to send a step to the smartphone, for example. We propose energy harvesting for photovoltaic, thermoelectric generator as well. In terms of sensors, we have accelerometers, gyroscopes, humidity, pressure, UV, a lot of sensors. In home automation, the key enablers of this market are clearly power saving. It's one very important item, also with comfort, better comfort, better power saving. Wireless connectivity is very important to deploy this easily in the house. Security is really important not to be hacked, you know, hacking by the neighbors, by whoever. Touch screen is important as well. Touch screen requires good eyes and clean hands. And it needs now to be completed, complemented with natural language, speech recognition. It's more convenient. Speech recognition is coming now. It's already very popular in consumer electronics and smartphones. A lot of TV propose now the speech recognition. Just to give you a little history, 47 years ago Stanley Kubrick in Space Odyssey was proposing and surprising everybody, all the fans, with this ship that you could command with the voice. There was natural language with the space ships. Now the technology, the brick are available for you to develop such possibility. So, excuse me, back one slide. So there are two ways of doing speech recognition. Either it is connected to the website, to the cloud and then you get speech recognition on the cloud. Unlimited vocabulary, unlimited language, so it could be English, Spanish, whatever. But you need a high speed connection. Another way is to do it locally. What you see on consumer electronics more and more is to have the microphone on the remote control, Bluetooth low energy link, speak to your remote control to get internet browsing, YouTube, video, daily motion, whatever. So this is coming in consumer electronics. And if you do connected objects, smartwatch or bracelet, whatever, you can imagine to have a microphone and speak your command to the object. But this is very short distance. In this case you need only one microphone, low power microcontroller, Bluetooth smart and power management IC. Is Bluetooth smart suitable for audio, Bluetooth low energy? You don't think so. It's not suitable at all. But we have a demo here. We show that it's possible and it's good. What we do is it's just a voice command. It's not high-fi audio. It's not audio streaming. And we demonstrate that we can transfer some voice command with 10 millisecond latency and very high quality enough to be recognized. So 64 kbit per second ADPCM encoding. We provide the software bricks for that as well. So if you want to go and develop such application, we have all the bricks for hardware and the software free of charge for you. There is a license of course. Any question about that? So this is the way we do for short distance voice command. We have good quality, 10 millisecond latency, 64 kbit per second. On the transmitter we do connectivity to the microphone. Compression cut into 20 bytes and send by Bluetooth low energy with very low consumption. For longer distance, this is what happens in the movie. You speak and the microphone is not here. For longer distance, then it's more complex because there's a lot of noise, surrounding noise. So how to do? We propose some software that does sound localization. So the microphone is able to localize you in the room and then to focus the attention of the microphone on you when you speak. So removing the surrounding noise. So the hardware bricks we propose is men's microphone, but then you need four men's microphones to perform this localization and beam forming. Higher performance microcontroller, speaker amplifier, why not, and connectivity. For software, we propose the microphone driver, sound localization software free of charge, microphone zooming or beam forming and noise constellation again. It's software, we don't sell software, we sell hardware, we sell components. So this is how the long distance voice commands work. First of all, sound localization, then beam forming, change the carduid by software, then remove the ambient noise and finally recognize the speech. So for recognizing the speech, you can do it also locally on the STM32 microcontroller. It's possible, you don't need high speed internet connection, but in this case, it's much more constrained like limited vocabulary, only one language. Okay, so it's okay for Espresso machine, for example. You say Restreto, Longo, I don't know, Espresso. Three words, one language. This is okay for speech recognition locally on the microcontroller. For this, we don't have our own software. So you have to deal with external company, external partners who are going to sell this speech recognition software to you. But we facilitate all this by providing all the other software bricks free of charge. Okay, so the two kinds of, oops, sorry, I missed my conclusion. So this is the end of the session. I would like to wrap up with these key messages. ST offers all the bricks for you to develop innovative speech recognition and other wireless sensor application for home automation. We have low power security. We have high performance Cortex-M, MCU. We have sensors, analog power management, actuators, and we have wireless transceivers and modules. All the hardware bricks. But also we bring some added value because we have all these bricks. We also develop the software. I bring it to you free of charge for you to go faster to the market. And time to market is margin for you. We understand that and we want to help you for that. Fast time to market is also done by our easy prototyping boards, a nuclear board, expansion boards, together with our algorithm and expertise. I thank you very much for listening to me. I hope it was interesting for you. If you want to know more about that, please meet me on the demo, on the booth. I will show you everything about audio, sensors, analog, RF that goes around the microcontroller. Thank you very much.