 Hello everyone, my name is Enrico Mazzoni and welcome to ST Microelectronics, both at this fantastic Mobile World Congress 2023. So ST Microelectronics is all over the mobile world? Yeah, we are a semiconductor company, but we propose you many applications and use cases. You can have a look at our demos and pods, they are brilliant and fantastic, introducing many novelties this year, active in various segments and parts. For an example, you have also the novelties about USB, USB communication, 360 degrees contactless USB. Please welcome. Hi, hello, so what are you showing here? So we are showing 360 degree rotation, contactless USB. So if you can come close to the microphone, what are these little devices here? Yeah, just one second. Switch to the main. So what we are having here is the ST60H3. This is a rest receiver operating at 60GHz. It creates a point-to-point link for USB 2 connection. Point-to-point? Yeah, it's point-to-point. So it's using kind of like host and slave at the same time? Yeah, kind of, yeah, yeah, that's right. And so up to 480Mbps, point-to-point connection, low power, like in the range of 100mW. Here is a cool application. So here we have a USB 2 camera, which is connected to a one ST60H3. And the next ST60H3 is located on the other side. So this is a standard PC with no USB driver update whatsoever. And this one is connected here. It recognizes immediately the camera. And now I can rotate. We can rotate and we can see that the video is transmitted. And the latency is very low. It's a cable-like behavior basically. Cable-like behavior? Yeah. So this is useful for many different things? Yeah, this is useful for many applications, for some testing, where you don't need to plug it and plug a USB connector for example. So it's very useful for wearables. But now there are more and more wearables having a wireless charging, as my colleagues are demonstrating here. Also people want to add data wireless connectivity to it. And this is a great product for this. Quite affordable USB connectivity, 400Mbps low power. And what's happening there? So here we are demonstrating a secure link. So typically the PC will be uploading a secure firmware. So we can flash application B. And data will be prepared. And the encrypted application is uploaded from the PC to the target through this link here. And you can see there is wireless connectivity here happening. Then the encrypted application is downloaded on the target. It is transferred, it is decrypted. And then a secure link is established. You can configure in a secure manner your target. Are you planning to do 5Gbps or even higher USB specs? Yeah, we do. USB 2 is ST60A3. Gigabit per second is ST60A2, which we are typically using for some contactless Ethernet applications. Cool, thanks a lot. Thank you. Let's have a look and continue with the IoT connectivity. I will bring you with me. Here you can find the latest technologies about cellular IoT and GNSS. We had just launched the news byte last Thursday about the ST87M01. Coming with the reliable and robust NB IoT connectivity to release 15 of the 3GPP combined with the GNSS and localization opportunity of the GNSS integrated inside the module with the option of the ST secure element and also the SME embedded, offering very robust, compact and low-power features. It addresses several IoT applications, widespread IoT, ranging from the smart cities with the smart lighting application to the smart metering domain, especially water meter and gas meter, battery power operated devices, but also heat meter and electric meter, ranging to the smart industry domain with predicting maintenance and other use cases and scenarios, and lending benefitting from the GNSS localization to the asset tracking solution and option. This solution can address various application options. It can come in the very simple flavor with the module itself combined with sensor and actuators for very simple IoT application and adding an external MCU it can also address medium to high complexity IoT scenario. Very, very new branded devices. What is the GNSS? Yeah, the GNSS is a Global Navigation Satellite System so it can offer you through the satellite data retriever the precise location with high accuracy and low-power features because it is integrated inside the SOC. Is it alternative to GPS? No, it is part of the GPS. It is a more combined and optimized solution because it adopts both Galileo and GPS constellation to have a very accurate and precise localization. Galileo is the European one? Galileo is the European one and GPS is the US one so it is a Western-based solution. Nice. And the NBIOT is a new trend. It just works. Is it using cellular? Yeah, the NBIOT is a very robust and secure connection because it is based on the mobile network operator infrastructure so it is part of the 5G family, the very new LTE family very optimized in power consumption feature. Cool. And we see here those little parts, is it in here? Yeah, this is the evaluation board as you can see very simple and very effective evaluation board with Arduino pin connectors for developers. It is fully programmable and the engineers can like it very much and can use it a lot. And here on the screen you show all the different connectors. Yeah, you can see the constituting part of the evaluation board. The module itself is in the center. Then you have the SMA connectors for antenna connectivity. Then you have the plastic holder for the SIM. Then you have the USB. You can connect both powering and sending data to the UART connection to the module itself and then you have the Arduino connection for the developers and software and firmware guys having the ends on this device. What kind of ARM core is it? Yeah, this is an interesting platform because it comes with three separate cores. So you have the Cortex M4, the DSP part and then you have the security part with the RISC 5. So it is very, it is an IoT platform itself. So you're combining tri-cluster, you have two ARM and one RISC 5? Exactly, indeed. It is very interesting because you have a possibility to embed also simple customer application and so you can customize the device and well address different use cases. Very nice. Is it ready for mass production? Yeah, it will land in mass production in Q3 this year. So just be patient a little bit more and you will have the solution in your end. This could be one of the most popular RISC 5 chips in the world. Yeah, this is maybe the primary solution inside this domain to adopt this kind of idea and novelty. So we are very proud of this insertion in the production. What did you need to use the RISC 5 for the secure part? Yeah, it is an ancient feature for the security parts that nowadays also for geopolitical reasons and interest in the security. Also for the Cyber Security Act of the European Union to be landed. It will be an IoT platform with security, additional security features landing in the European, US and worldwide application scenarios. So this could potentially be the new standard for country-wide IoT? Exactly, indeed. We are part of several groups also for the standardization in security so we are driving the working group about this feature and this will be the enabler, one of the first products to disclose this feature. Nice, that's cool. Can we check some of your colleagues? Yeah, sure. Yeah, let's move around. We will have another very hot topic, like connected artificial intelligence. I will introduce you to my colleague about the connected AJI. Please. Hi. Hi. So what are you showing here? So actually here we are showing an example of the application that can reach by embedding AI machining on the edge. So actually we are running... Can you come close to the microphone? Yeah, sure. So actually we are running AI on the edge in order to reach best performances in a lot of applications to create new features. For example, here we are showing how to create predictive maintenance solution by monitoring the vibration of the pump and allowing the user to take his solution, to put it on another system and to learn, again, the behavior of the equivalent in order to venture the model. That's what we call on-device learning. So to do so, we have a tool called Nano-AJI Studio that you can see on your right. This tool allows the user to create from scratch in a matter of minutes the AI algorithm for automated detection, classification or regression and to cover every kind of use cases. We are sensor agnostic and only takes few steps to create an optimized machine learning model. So you can see that the studio is doing most of the job to ease developer life and we are reaching pretty huge performance, almost 100% accuracy, with a really low footprint. So these solutions are really optimized for SM32 microcontrollers and allows to create an immediate kind of use cases. We are monitoring here vibration, but we can work with temperature, current, pressure, audio images. So the number of use cases is limited for that. This Nano-AJI Studio, what is that about? This is the tool that will speed up the development process for AI solution. So it's optimized for microcontrollers, for SM32 microcontrollers and allows every developer to create machine learning solutions. I've got a comment here. Please ask how to embed the application along with the OS. Is there any manual to set up? Actually, the embedding process is quite easy because the studio is generating .h and .a. So these files can be implemented to the project and it's just like calling a function. So the AI is here allowing it's a technique to reach best performances, to reach new features, to develop new features. So you can just create many new projects and call the function as you used to do. What are the skills required to program for this? Actually, if you want to use Nano-AJI Studio, you don't have to get advanced skills for AI or data science because the tool does a job for you. If you come and watch, actually the tool will benchmark thousands of combination of machine learning models and parameters and preprocessing algorithms. So it will define the best combination for your application. Here we can see for this demo that we are using three scenario preprocessing bricks and one classical machine learning model. So the studio will tell you which model and which preprocessing you should use but you have the right to change if you want. So this is giving you all the tools that you need to develop your application so you don't have to have any skills just to embed the library in your portfolio. This is exciting, this is providing many new potential implementations. Exactly. Thanks a lot. ST is very active also in the imaging part so let's have a look at these three other options. We have the current cabin imaging demo pod. Nice to introduce my colleague. Hello. Hello. Lutastaman, I'm a technical marketing for our imaging automotive products. Are you interested in automotive? Yeah. What do we see here? What is this device? This is our EVK together with Illuminator here. So this application is for in-cabin monitoring and user experience features. This is a new innovative 5 megapixel sensor which is unique in the world because it is capable of doing rolling shutter and global shutter mode at the same time or let's say like frame-to-frame you can switch between HDR RGB pictures and NIR here with the illumination. You see the NIR, the NIR infrared 940 nanometers illumination. So that's 15 frames per second of each 15 frames. It is just a demo. The sensor can do up to 60 FPS in full resolution. 30 each for example in full resolution, yes. How does it switch? It switched. There is a firmware inside. There is a little controller in the sensor which is switching then frame-to-frame so you can pre-configure the sensor in four different configurations and then you can set a sequence and then switch between the context and the sequence and you switch frame-to-frame. So you make one picture in this mode and one picture in that mode and then you stream the data out to an ECU and then they can do some kind of orientation like this when you go here on this side you can see that it recognizes my face where I'm looking. If my eyes are open and this is a safety application like the European Commission there is this NCAP regulation that every car in 2025 has to have... No, it's only working here on my side because it is configured. This is the driver side. So from 2025 every new car in Europe has to have a driver monitoring system. That soon? That soon. Yes, yes. So we are very busy with OEMs and with the TO1s to implement that in cars. We are on a good way to get the market lead also here with our products. We have two products. One is only focusing on this application with NIA infrared and then this product is a wide field of view for the whole cabin and then you can have all applications in the human and in the machine vision side. I like watching these YouTube videos with the Uber drivers. They film themselves and they show what's happened, but this is like filming inside the car. Yes. In full color here. Yes, HDR. Security. Exactly. So this is machine vision. Normally you don't see that. The machine is seeing that and then gives you information or sends you information. It makes beep, beep, beep, beep. Or maybe it peeps this. So for the DMS it's only like the drowsiness and the inattention on the driver. So when you have like... Exactly. When you... There's another mode of this demo but it's not enabled right now. So when you have the mobile in your hand it recognizes and then it would also like beep and say please don't call during driving. Don't play video games on the dashboard. During driving in your Tesla. Yes, exactly. So here this is also the new trend in the automotive, like the software defined vehicle. You bring more features what you know from the smartphones into the car and this is why you need human vision and nice HDR video content also to the people in the car or to stream that making zoom calls or teams during the driving. I need this function where if I make a joke that's not very good with my guests you should just park the car to the side of the road and stop... Yeah, that's the beauty of the software defined car you can implement everything what you want. Because the big problem is distraction. It's really a lot of accidents. Yes, exactly. So this is why here for example this is the end of the question again. So it knows where I'm looking. It makes the eye tracking and then you can say now I'm looking on the side window now I'm looking on the... on the video which is in the center display and now I'm focusing on the street. So with this sensor we are really cutting edge and we can also do some processing on the sensor itself to make out of this this is actually like an RGB NIR 4x4 color filter array which is not normal for normal cameras normally you have only RGB and the standard ISPs are not able to deal with this so you need to process this or pre-process this and that the standard ISPs are capable to use it. What we can do here with this sensor we can do this pre-processing on the sensor and we're the only ones in the world which are capable of doing this. Nice. Yes. Alright, thanks a lot. You're welcome. Cool. Let's have a look also to the other solution like smart cameras and photonic sensor. Hi. Hi. So, yes. Here it's a global photos camera so we have three of my colleagues with in-carry solution it's a consumer and digital solution so we have a lot of type of application possible like the first one it's a face recognition so for that we work with Trinamix The mic. Okay. We work with Trinamix in the partnership with Trinamix for the face recognition so we provide the camera they provide the algorithm and the software solution so we show today we are able to have a great level of securities for mobile payment because we are a field of security certification for that and we show we are able to go under the screen with this camera we can go under the screen that's a main point for us to do that because we are a great sensitivity with our sensor camera in higher wavelengths So this is kind of like face unlock face authentication for payments for payments for screen for all you want for smartphone unlock and so on Some of these phones are using kind of like LiDAR or something like that that recognizes the shape of your face and everything Is that required sometimes to have extra level of for LiDAR? You mean? So you cannot unlock it with a photo with a person No, no, no With our camera and the algorithm from Dynamics we are able to the difference between a mask a real person a mask and a real person because they see the skin texture skin texture and movement and skin movement So we can know that he's a real person or not this real person So that's the main point Does it take a second or how long does it take to check that the person is moving and everything? One or two seconds Not more, not more Yes Is this to be implemented inside the phones or inside what? It's for it to be implemented inside the phone and that now available because we can go under the screen so we don't any more of the notch or a smaller point for the camera we can go under the screen for this camera What's the difference in quality for a selfie camera that's behind the screen compared to a little dot? Oh, yeah The main difference is we need great sensitivities because to go through the LED or the screen we don't have a lot of transmittance so we need to have a great sensitivity in higher visibility and we have that Cool Is this available now? Yeah But not yet in the phone Not yet in the phone but we're in the three limit booth if you have time to go and we have a demonstration with the same camera under the screen under the screen there we can see that we can use it that functional things that are available in the demonstration and it says a number of pixels megapixels and stuff like that It's a 1.5 megapixel solution in global shooters All right So that will go behind the display and then you can have a different selfie lens also You can have one for selfie photos another one just for autotikation Yeah, for autotikation we need to have an higher, higher source because we need to have real abilities so we need to we need to go when there are dark scenes we need to have a face identification too so we need infrared sources That's the main point So yes, today the main solution, use it Today we have that on the phone we have higher sources on the phone It's pretty nice that you can use your phone to pay contact less and you don't need to put a pin code or anything It knows it's you Yeah, it knows it's you too Exactly It's exactly the main point It's the main point Nice All right And there's an encryption and maybe octi or something you use encryption on the chip that goes behind that Yeah, we can go to this type of application All right, cool Thanks a lot Thank you Can we continue the tour? Yeah, sure We can go back to the USB because we do have other technologies and interest on the USB topics Yeah Hi Hello How are you? Please introduce yourself Yes My name is Christian Bella I'm a product marketing manager for power supply and USB PD products So nice This one So what do we see here? We see here a 140 watt charger which is a very compact one The first one in the world to get the USB PD 3.1 EPR certification for 28 volt 5 amp outputs This is really compact and lightweight So it is used Gallium nitride Correct It uses gallium nitride transistor of latest generation and digital control So the combination of digital control and gallium nitride transistor allows the dimension to be very small and the magnetic components to be very tiny and lightweight because the system operates at high frequency 300 kilohertz As you can see the efficiency of the system is above 94 percent and this is 2 percent higher than average chargers Also the dimensions are three times smaller than a solution of the same kind but using standard transistor What do we see here at the table? The next limb and the magnetic Those are the transformer done by, yes And those are done by our partner LG Inotec who helped out to create this prototype here which is in for factor solution for USB PD charger What's here? The EB loan It shows what's inside a charger That's correct Inside the charger this is the circuitry showing the ST1 HP digital controller and the master gun gallium nitride half bridge power stage Those are the two main ingredients for this solution which is an active clamp flyback converter and PFC controller in front So the total solution is made using ST components and LG Inotec magnetics together with Vort electronic magnetics So it's a total solution and this is what's inside this charger here Nice, it's a lot of power 140 watts It's kind of amazing that chargers can be so compact and reliable, safe That's correct In fact we got all the certification about compliers with USB standard and it's also safe as we said This 140 watt is enough to charge a laptop a smartphone with very fast charge You can charge a smartphone in less than 20 minutes and a laptop in less than 30 minutes And you can also use it to charge an e-bike for example or a power tool, like a drill or even a vacuum cleaner a portable vacuum cleaner So it's a really universal charger And it said the world's first 600 volt Is that what it said on the screen up there? Actually it's the world's first to get the USB PD 3.1 certification and we were also the world's first to introduce the master gun solution which is a 600 volt 600 volt half bridge power stage using Gallium Nitrise Transistor of second generation And this technology can be made available to every charger manufacturer in the world That's correct Our technology is available for any manufacturer to use our technology to take advantage of our technology and create products And if you replace 1 billion chargers you'll save 20 minutes you'll do a lot of savings Yes, yes So if we could replace our charger in charge with our solution we can save a million tons of raw materials and energy saving for 3.5 billion tons of CO2 So it's an amazing amount of savings So actually it would be a good idea even though it sounds like a lot of waste to change all the chargers in the world but in terms of how much power you would save in efficiency it's worth it for the environment That's correct And in fact the USB-C is the connector which is chosen by the EU the one which replaces all the USB connectors So nowadays USB-C is the universal connector for all the devices And now the new USB standards are trying to make it clear how much bandwidth and how much what throughput power does so it will be more clear what cable is supported what charger is supported And that's also very good, no? That's correct On the cable you have the ability to know how much power you can deliver and so with 140W you can really charge a large variety of products I like my power bank it's very interesting because you connect and it tells you how long it's left how many hours, how many minutes or maybe it would tell you how long it will finish charge All this kind of information could also be in displays and stuff But that's something else, right? Yeah, that's really useful and convenient for the end user Thank you very much We can end up our tour with the very last one pod in front of the desk Please, let me introduce you the self-learning IE sensor pod that we are presenting with the ISPU product Hi Hello What do you show here? So here we are showing our new IMU which is an accelerometer and a gyroscope and inside the same package we also integrated a risk core where in this demo we are running unsupervised learning for fitness tracking So we are able to learn some movements and then detect them What am I seeing here on the table? So on the table you can see the sensor which is the small part right here Then you can see the app where we can see what kind of exercise you are performing and how many repetition counts you have And also for the sensor there we have an evaluation tool with a wristband And you say risk? What's the views in there? It's completely one made by ST So is it an ARM CPU? No, it's a completely... the architecture is made by ST It's not an ARM Alright So this is useful for... What does it do? Self-learning Yes, so as I said for fitness tracking for example but also you can run any other algorithms you like We provide a tool chain for it so you can program it in C language you can run your custom C processing algorithms any of you like Cool, alright Thanks a lot Thank you Alright Alright, thank you so much for passing by at ST and we always welcome you for new developments and new excitement and things to make life augmented as our motto is saying Thanks a lot Thanks for the tour And it's been a good show right? A lot of people passing by We are very enthousiast of the people passing by A lot of people A lot of interaction, customer new possibilities, new technologies It has been a great 2023 MWC show Yeah