.jpg)
.jpg)
.jpg)
.jpg)
Good morning. Ladies and gentlemen, I'm so grateful to be here to see all of you. You know, yes, today is the IIoP Partner Conference, but also is a five anniversary of Aleventai Turkey. Now I established Aleventai Turkey since the two thousand nineteen, so that means today is five years old. Before, to come to the event, I tried to thinking about why I run the Turkish in the past five years. I know we are foreigner. Right? Anytime we come to the new country, we try to learn your language, like, hello, like, nice to meet you, something like this. So I'm thinking what the Turkish I've learned in the past five years. Eventually, only one word showing in my mind is Virasem. So anytime we come here and enjoy the business with all our partners in the past five years. In the past five years, at the event here, we have a very strong commitment and establish our organization in the territory. Now we have more than thirty people in the Istanbul, in Busan, and in Ankara. Okay? And you know that seventy percent of our team is for the support and the service. Why? Because we would like to empower our local organization to support our friends, support our partner, support our customer, your business in this market. So that's how we did in the past five years. And now what we will do in the coming futures. The first one, for the Turkey, we have very strategic positions in the global, and we connect. We are the heart of the world. Right? So we connect. The euro connect with Asia. So at the moment, we like to establish at the moment that Turkey as a the Turkish and our business as service will cover all the Middle East area. The second, we're talking about the IoT business. So what's the IoT business? So we are doing the transformation, moving to a solution business by different verticals. So we will offer the hard way and the soft way combined together, plus your domain knowledge for different verticals. So we are invest the strong local solution team. We call it domain focus, the integration team in the Istanbul to support all of our partners to stronger your solution than in the business in the country. And the last one is talking about co partnership. So we're working with our partner for the different vertical, like you. So we invite you to come here to join our partner conference. We have some partners focused on the the general hardware selling. We have the GCP. We have the FCP. But we have more partners focus on the different verticals. So we're working with them to co create a solution by the different vertical Okay. Vermaat, I learned the ninety. That's that's all I know. But different from Vincent, he took five years to learn one word. I took five years to memorize. And, it's a pleasure to be here in Turkey again and to see, you know, these familiar faces that I've met, in the past few months. Today, to start off the conference, I would like to share a little bit about, Advantech and how we foresee, AIOT plus edge computing, will be our strategy for the future use. Okay? So first, I think most people here already, know Advantech or know Advantech mostly from the hardware world. And yes, for the past forty years, Advantech has been, consistently working to grow the business in the, local PC and I would say more in the hardware world. And that's not going to change that much because it's the core business and it's also part of the DNA of Atlantic. And in terms of numbers, the market size, it shouldn't change either because here, as you can see, the potential of, hardware, around the world by twenty twenty five, it will jump from the one hundred and seventy five million to two hundred and fifty million. And here in this table, you can see from different, sectors, there is, several opportunities. For us, going forward, we will focus on the fast growing sectors, which I will explain later on. Okay? So from hardware perspective, there's no change. But what about, the future? Vincent mentioned that if we only consider hardware sales, it will be more and more challenging because it's basically, comparison of, hardware, requirements, configuration, the technical specs, and then price. Unless there's not much secret on that. But if we consider going forward, how can we include AI into our solutions? Right? So I think everybody hear a lot about Edge AI, edge AIOT, but, several times we we kind of, get stuck on how to execute or how to move forward. Right? So we all know the benefits of, AI. We all know the the the the the process cost. But, there are two things that they usually block, the implementation of the products. But first, I will say the first challenge is how to improve in our, solutions and our proposals the maintenance and update of, our products and solutions in general. Right? Because when we're selling, solutions, we're not just doing the presale. Right? We need to consider that, there's, usually high investments involved, so we need to think about the future. Right? We're talking about, solutions that should, remain for five, ten, or even four years. So one thing to consider is always consider maintenance and update issues in our solutions. Second, is the different restraint. I think this is, one of the feedbacks that a lot of customers, give to Advante. Right? So what about, security issues? What about compatibility issues? Right? This is also one of the key, factors that is blocking us from evolving faster or accelerating the execution of the crime. Right? So I'm I'm saying all this. Why? Because when we talk about, first the market in terms of hardware, we can see numbers. Right? Then when we talk about AI, we can see a lot of, opportunities, but there are certain restraints. So what I'm trying to show here, that's also why where Atlantic is trying to, evolve into is to become, instead of pure hardware provider, is to become tech and service providers. Right? And why is that? Because if you remember the first slide, we talked purely about, hardware, we're talking about, two hundred fifty million market. Right? But if we consider, hardware plus software plus the service market, we're talking about double, almost a double of the the the market size, at around five hundred million. Right? And these are all, it's not coming from Advanced. These are informations from Garner and from other, other third party companies. Atlanta. Second is going forward, we can see that the ratio of pure hardware, providers will be migrating to these so called, TSS, tech service providers. But from twenty five percent, probably, reaching even sixty percent of tech providers. It transitioned from pure hardware, solution that that pure hardware providers to solutions providers. Right? And third, this is very important, is providing low code or low code type of, solutions in order to speed up implementation. Then, what is that, how does Advantech come into this strategy? Given that, we have identified that going forward, we need to start, focusing in, the development of different sectors. Right? So there are several, main sectors, but for Advantech, we have identified that going forward, our product and, software or total solution will be mainly focused on these verticals for manufacturing, transportation, energy, retail, hospital, and I health care related sectors. So the company were, starting to to to, see this, sector driven transformation. So at this stage, we have split our business into eight main sectors. And for today, we will mainly focus on the number three and number four, which will be industrial, edge solutions and also industrial automation solutions in general. Okay. So what what we foresee because at Verintech, we are one government. And as I mentioned, we're very strong in the hardware, hardware part, but we cannot, we cannot embrace the whole market. Right? That's why we need to create this ecosystem and we need the help of all our partners. We have our diamond, our gold, our silver, and, bronze sponsors, as well as several other customers, we need our partners to help us in building and pushing this ecosystem going forward. And as you can see, from Advantech, we are positioned, I would say, in the middle as a facilitator of this kind of, activity. Right? So in the left, we have all the leading, companies in technology worldwide. Even today, we have, the honor of inviting our partner NVIDIA to provide us also, a speech, about, market trends and, AI. And I'm back in the middle then we accelerate these kind of, solutions through the different markets. That's why, also Vincent mentioned, that's also one of the main reasons why Atlantic invests so much in our local, structure. So building up our local teams in order to serve the market needs, low instead of, from Taiwan, which is very strong. And then just, this this is the, three layers of, Advantech's strategic framework. So because in Advantech we have several different types of hardware, we also have our, Wise Pass solution. So it's, I know it's it's hard sometimes it's hard to remember all the products or all the solutions that Advant is providing. For me, it's really hard to memorize everything. But I think, I hope like, the sharing today, is that all of you can remember the architecture. So remember the architecture, you can identify the opportunities and how Advantech can co work with our partners to provide the right, solution to the customer. Right? So three three layers. The bottom layer, we we have named it Platform. So basically what it is, is all the different types of harbors that Atlantic have, from sensors, connectivity, gateways, Edge computing devices, and so forth. So at the very bottom layer, we're talking about platform. If we jump to the top layer, then it comes to the different sectors that we have identified and that we will be focusing on. Right? So in the middle, we call it the orchestration. So how to link the hardware plus the different types of solutions and applications in different centers. So based on that, we have different, software solutions from our side, which is, EdgeSeq three sixty plus our WISE IoT platforms, which are the, orchestrators or the facilitators in the deployment process of the the the project. Okay? So it's just simple to remember. The architecture is a framework. Right? Bottom line of the platform, at the top, we identify the different sectors that we want to approach, and then we have, different types of softwares from Advantech or from our partners in order to provide a solid solution to our cloud. This is a more detailed vision based on the framework that I just described. Right? So as you can see from the very bottom layer, different devices, sensing devices, gateways, edge computers, IPCs, etcetera. Right? If we go on to the very top, it's all the different, sectors that we have identified. So this is a very general, architecture of Advantech Solutions. But, based on different, sectors, we can have we also have it, but due to time restrictions, I'm not sharing each one of them. But we have the same architecture that is focused in, let's say, smart manufacturing, transportation, eye health care, and so forth. Right? So this is just, so that everyone can, remember whenever you go to the customer side and you see an opportunity or in the demand demand, then we can easily identify where Advantech can help our partners in the business. Okay? And last, wrapping up, that's the vision that, Advantech has, changed in the past, I would say, fifteen years from pure hardware, provider, now where we're coming and we look forward to the future with this new vision, which is enabling an intelligent identity, not only in Turkey, but I am a visionary, illuminating galaxies to witness the birth of stars. And sharpening our understanding of extreme weather events. I am a helper guiding the blind through a crowded world. I was making more running to the store. And giving voice to those who cannot speak. And do not make me laugh. Hello. I am a transformer. Harnessing gravity to store renewable power and paving the way towards unlimited clean energy for us all. I am a trainer, teaching robots to assist, To watch out for danger. And help save lives. I am a healer providing a new generation of cures and new levels of patient care. That I am allergic to penicillin. Is it still okay to take the medications? Definitely. These antibiotics don't contain penicillin, so it's perfectly safe for you to take them. I am a navigator generating virtual scenarios to let us safely explore the real world and understand every decision. I even help write the script. Breathe life into words. I am AI. Brought to life by video, deep learning, and brilliant minds everywhere. Good morning, everyone. My name is Christina, and, thank you for advocating me to, give you the speech notes about AI. I hope, everybody will get something out from my presentations. So, today, I'm gonna talk to you firstly about the, the marketing, market opportunity of AI at the edge, and then we introduce the overview of our objective platform, and then I'll discuss about the software of how, NVIDIA is offering to everyone. And then I'll wrap up the, presentation by offering you some of the successful stories around the globe. So let's have a look at the, the mass market of the AI. Why is AI needed? So, apparently, we have a network video recorded, about two hundred million and thousand eighty streams, And, also, there are one trillion machine AI on the, on the field. And, also, home service, robots is all the, artificial intelligence that is well used. And it's expected to, be at the size of seventeen point eight billion by twenty twenty six. So it's showing a very positive growth rate at about that thirty percent. That would work. That is why AI is important for everyone. Yeah. I would like to discuss with you about Jess because I think many of, many many of the other body in public still think that NVIDIA is a gaming company until about a couple years ago when, AI booting starts. And, I wanted to show you that, our Jetson platform. So it's not just a GPU card that you are expecting. It's actually constantly, a module where it is, it's a step on module. And what is so special about Jetson is that it comes with a soft it's a software defined AM platform. What does it mean? You buy a Jetson module, you get all this software for free. That's how we are different from our, competitors. That means we offer the, the Jetpack SDK. And what is Jetpack? It's like, iPhone iOS system. You get updates, constantly, and, of course, this is for free. And the updates able to do further optimize and improve the performances of the the adjuster modules. And on the other hand, we also offers different type of the SDK as soon as different type of verticals, which I will come into the detail in the following slides. And, and another, I wanted to mention is that we have, we are, drive to, improve our ecosystems. So it's like once the customer purchase the module, it's not like they could've left alone not knowing, not going to get their cameras or software to help them to further optimize their applications. We have sensors, cameras, and then also robot computers. For example, that's the, Advante, the big box that's that's, displayed on it, and there are demos outside as well. And, all these, ecosystem partners are verified and certified by, the media. And, of course, Advantech has its own vendor tools that are also certified as well. So you are not left alone when a customer buys the modules and then not getting all the accessories to build your models. Oh, I just look there. Alright. This is showing you that all the, verticals that we'll be touching also so you're not just covering everything, and, these are all the, customer key customers that, we have. And at the moment, we have about thousand three hundred k developers. That's not the media people. It's actually the developers of the end customers. Yeah. Because somebody asked me the question of how what it means. And then, also we have ten thousand plus, end customers worldwide. And so how does, the media assessment solves a problem of AI development? So as I mentioned, in the previous slide that we have it's a software defined platform. So the customers are able to develop their, program, they I mean, application based on our, supported SDKs. And also, we have, a comprehensive hardware and software, that's available for, to depends on different needs. And also now we have a a really well defined ecosystems that allows the customer to, to do, like, a one stop shopping. They can now find whatever parts they need within the ecosystems. And like I said, Jetson is not an IC. It's actually assistant or module. So it's got, like, SoC, memory, powers, management, IO, and extension just full on it. And this slide is telling you the softwares that we offer. So we understand that, training the, AI models is taking enormous time for the, developers. And what we are able to offer is that we have a pretrained models library. So what does it mean? Like, imagine you have to write a thesis from scratch. It's very difficult. So you have to you you can go to the library to borrow some books and coming out with your own ideas. So this is more or less the set of, scenarios. So the treatment models that we have in the media libraries is about eighty percent. I think at least eighty five percent plus, accuracy. So the developer can use some effort to optimize to suit their needs. And we have the tile tool kit. This is also helping the customer to, train their AI models. And we have different SDKs. For example, ISAC is for the ROS, robotic, development, a D3 for computer, envisions, and revise for languages. And, we also have the Jetpack. It's for the hardware, the Jetpack modules optimization, which I can discuss that in the in my following slides. So this is a return model that's available for the, for the part for the, the Jetset customers to use. And you can see it covers flights of verticals there, smart city, pocket safety, retail, health care, robotics. So it's at least minimum of eighty percent accuracy, and this is available for the, customer to use to, they can freely use it without any problem, with the money or licenses or something like that. This is all for free for the customers. And then they can be it can be easily deployed and then transfer the the, with customers to start sheets. And, of course, customer can use their own data as well and to integrate with it to make it more comprehensive. So, yeah. Let me show my stuff. Alright. This is, just to show you that on the right hand side is a video frame, people net model. You can see it's got obvious that are it's about four or five times better than the, the hand side where there's people detected. But, I think a a more important message to stem across today is that, I often have a customer telling me things like, why do I need AI? It's so expensive. I don't need it because it's so much money I have to put it in. It actually depends because I have to say you know, all this customer say, you know, this one doesn't really define anybody. Nobody's gonna buy this. Why you say that? Because it actually depends on your application use. For example, if you I was in the Zulu Center the other days, so it was quite obvious that seven o'clock was very, very busy compared to twelve twelve PM, one PM. And for Zoho Center, if they buy, if they want to determine how many people's at that shopping mall and to deploy how many service staff or cleaning staff to be at the shopping mall, they can have something like this because you don't have to define everybody in the, in the zone center. You just have to know the trend. For example, at one PM on the weekdays, it's it's not so many people. They're learning from the AI. It doesn't it just shows that, you know, a few people's there. And then in comparison, that's seven PM with a lot of people. And this model will be, this this one will be a a a right fit because it doesn't need to be so precise like the right hand side. You only need to have, like, you know, a rough number to be able to pull the floor stuff. On the right hand side, however, for example, for the airport, where you have to chase down somebody from the in control or, thief or, like, a criminal, and then you have to have it, like, you know, almost defy the people, then I have to say airport is a suitable, security, civilians needed. So the right hand side will be, will will be right. So I'm just trying to say that, there are different uses for AI, and then it doesn't always have to be so high at, like, right hand side. The left hand side can be also very useful targeted for different market and verticals. It presses a lot. Right. So generating AI. So last year, since the chapter GBT was, was, available to the public and generating AI has been, a talk about, wildly. So I wanted to give you an understanding of what generative AI is. So the traditional AI is, tell me if this is a cat or if that's a dog, you know, as there are many cats, how many, there are many cats and then, you know, dogs, and they will just tell you which one is a cat, which is what's a dog. That's a traditional AI. So what is general AI? For example, I can input text or audio saying that, draw me, a ginger cat running across the tsunami. Then whatever I say in here, it will come out for, like, a movie or, you know, a cat running across the tsunami or a picture of a cat running across the tsunami. So anything you input can come up with a different context. It's creating a new thing. It's not like telling you cats, pseudonym. No. It's that I'm telling you to to creating a new context. That's what generated AI is. And there are different models at the moment. We have ROIAM, which is widely heard and visual language model installed. So Generated AI, why do we need to run on the edge? Firstly, because of cost, and also I have customer coming to me saying that the privacy of flying capability is more, concerned. And why why is that? Because if you because at the moment, we have making people upload the data from the cloud as, you know, as usual, practice. But there are concerns with if you put on the cloud, you know, don't we have to worry about any cap or something like that? There's always something you have to compromise when you put on the cloud. So if you put it on the edge, you can, of course, have it encrypted so it make it more secure. And then, if you if you have time during the break time, we have, MIC seven thirty three, a o, and also, that's running on the, generative AI context, and that's very interesting. You can have a look at this. It's right on the, outside of the right order. And yeah. So that's, something to consider when you approach the customers. And, just want to show you that we also offers the, like I said, we offer many software package for the, customer. And this is an MJ for the, So if you I think it's easier to look at it, you know, if you go outside. We have this demo right outside the, the in the demo room on the right hand side of seven three three. It's running a generated AI demo. So if you stand in front, it has a camera. It will tell you if it's a man, or a woman that was standing playing around with it. It has a a text telling, describing, what I was doing in front of the screen. So it's very interesting. Please take a look. So, here goes to the Jensen families. At the bottom side, it's gray shade. It's not because they're facing out. It's that the t s two is, unfortunately, facing out. But these are the legacy models. And, on the top, on the top is the, all ring is the current generations. And on the left hand side, we can, on the left side, you can see the entry level models for a lower, a lower performance needs. On the right is a higher performances. So, today at the demo showground, we have a MIC seven one one that's running on the Nano. You can take a look at it. It's, more an entry point. So if a customer doesn't need something that's so, pricey or too heavy duty, we can prescribe them the Nano over and over Nano as well. And for for the, generated AI that's running on the high performances because it's real time. So it's on the AGX order. So they are all outside, take you know, do take a look, you know, when you have time during the break time. Just wanna quickly, share with you that in earlier this year, we had this GTC, event in the US, and our CEO, Jensen Huang, gave a speech about the, gave a speech, and then he was announced with a new Jetson, agent Thor. So agent Thor is not intended to, replace, Orem. Orem is not intended to replace that. It's actually a different platform. So it's, it's actually equipped with the latest, NVIDIA Blackwell, excuse me, Blackwell architectures. And, obviously, this is very, very powerful. It's, like, eight times already given GPU performances. So this one is more or less for the, robotic related customers. For example, Boston Dynamics, I guess everybody have heard about them. And, the availability is by the second half of two thousand twenty five, and I can imagine if it's ready by Adubanketech and YV two thousand twenty six for less. So, I believe our Adubanket partner will keep you posted if their customers is interested, and, of course, it's very huge duty because it's for the, for the worldwide news. And global ones and still growing. Very good. So the software so we like I mentioned earlier, this is, Jetpack, and we have just released the six point zero version targeting for Orum. And then, we schedule for another release in the second half of the, two thousand twenty four, and this is for optimizing the hardware itself. And, this is something that our developer provides very much because we have, online forward. It's a four hour set of monitored and moderated by the, a media engineer. So if the developer has questions, they they are welcome to ask a question or interact with other, So, these are the six major industries which we've heard from you, and, I have specific picked some of the stories, to incorporate with Adabatic Test strategy. This one this one is a railway, it's a railway assistance, and then it's, AGX Xavier modules that's its core mounted owner. And it's measured the, the energies, consumptions, and then, determine, like, how it can be further improved. So, this has successfully helped the customer to, to, improve by, type of person that's what it says. And then, yeah, we hope to replicate this solution to other parts of the globe. And this one is a very interesting case at the, Yats Week in London, and I just wanna let you know all these case are already two to three years ago, that's why they're here because otherwise it's confidential. I'm not allowed to talk about the one that's going on right now. And this one is a case in London Gatwick. So basically, it measures how long this is for taking the customer to go on board and, and disembarking their flights, how long does it take to lower catering, onto the plane, and also on lower planes. And I think this can be something interesting for the Istanbul airport because it's a very busy airport, and London, yeah, since they have implemented this, solution, it has successfully saved, energy, c o two emissions, and also the, they are more efficient in, disembarking and loading on the, airplane. I think this one is also one very interesting that can be, interesting for the Tokyo market because we have so many fabrication that's made many many of the electronic products are made in Turkey. So, this one is the, a typical AI AOI, defect detections, and I think this, yeah, this can also be done by Jackson, and they can be, interested if you have any, if you know any of the Turkish factories who would like to do, implement this. This can be a a nice, application to think about. And we have, this is the Toyota case to determine if the, the, workers are complying with safety measurements. So this one can also be a very, interesting application to store in Tokyo because we have many manufacturers and clients in here. And it says it's about ninety eight percent real time error detection, so it can be quickly, prevented for, unnecessary, work harms because that would be more expensive for the cost for the, for the factories. And this is another industrial inspection as well. So you can see that, Jetson is widely used for the, for the, detecting the, the problems at the, at the plants. And this one is a very interesting one. As we know the, at the construction site traditionally, I actually don't know how to do it because, you know, they are digging there. How do you know if you this place has to be dug? And then tomorrow, you are not digging the same place, you you know, it's this is a very interesting case where they put the cameras and and the lasers of the, on the on the, exact it's tabulator, which is the equivalent is called. And so each day, it's recording what it has been construct, what it has been done, you know, like the progress, and then send a real time log to the, to the construction site program. So in this case, it's conflict free because it's we're not gonna have workers say, hey. I've I've done I've done this, you know. And then, all of the constructors, I know you didn't do this. This is prevent the, the problem of that. Finally, I wanna show share with you with this, waste, material processing, and I had a discussion with, our our, our, our, our, our, our, our, our, our tech team. So I've learned that at the moment that the rubbish the rubbish sorting is still by, manpower human in interview, and then which is the same in the API region as well. But something like this can be used for words, for fruit. For example, if you wanna pick there are many tasty fruits in, in Turkey. I was enjoying this fruiting when I come here. So this can be something like, you know, you can teach the machine to pick the orange. For example, this orange is very good. They can be sent to the market, and this orange is not so good. It can be sent for the juice packing. So this can be a very interesting case for you as well. So that's all for my presentation. Thank you very much for your time.