Infrared image sensors seem to be quite a big trend. We have various NBA rights, so it's hard to say, but we want to focus on cars, right? When we think about adding something to cars, the first thing that comes to mind is adding a camera, right? So, we are trying to drive autonomously by adding a camera. Actually, when you think about driving at night, sometimes even with our own eyes, we can't feel the distance well. Sometimes when a car in front is much bigger or smaller than I thought, I step on the brakes. It's because they are either too close or too far from what I imagined. So, just a camera is not enough for autonomous driving. People talk about adding lidar here. To put it simply, lidar quantifies the distance.

Isn't it strange? The image sensor actually doesn't give us the distance. When you think about it, it's just a 2D image. It's just us assuming the distance based on the shadows. So, if there is an object in front of us that is much larger or smaller than what we expected, we can misjudge the distance. In a car, even a 2 to 3-second misjudgment can lead to accidents. That's why now, there is something called LiDAR that people are working hard to measure distances. The issue here is that when you think of LiDAR, most people would think of something spinning around, right? You all know LiDAR like this. The problem with it is, first of all, it's ugly. That's a problem, but cars accelerate a lot. So, with all that acceleration, any device that has momentum...

It is also difficult to apply it stably to a car. Usually, image sensors and lidars are used together. I think the approach of using only lidar will disappear. There may still be approaches that only use lidar, but since we use image sensors and lidar together, there is also the difficult issue of calibration. In this image, there is a difficulty in accurately measuring the distance, so even if it measures around 10cm, when we drive, if you actually drive in Europe, you might think you will die. It's much harder than driving in Korea. However, the major automotive markets are Europe and the United States, and Europe cannot give up autonomous driving, so it's not 10cm, but they aim for an accuracy of 1cm or 2cm, making it even more difficult. Therefore, finally, calibration traps are needed, and if it breaks down once a year, calibration needs to be done again.

It's like, yeah, it's difficult. It's just a matter of level. When I tried to use the camera, there was this issue, and when I mixed in the LiDAR, it became even more challenging. But today, I drove for almost an hour and 40 minutes, and I couldn't see well in front. Also, as I went in and out of tunnels repeatedly, my vision suddenly got blocked for a moment, right? Still, since people have good instincts, they slightly reduce speed without any issues. Actually, even if you don't slow down, unless there's a risk of an accident, you just roughly assume others will drive similarly, right? In about 2 to 3 seconds of not being able to see, honestly, you just pass through. That's how I drive. But I don't think robots should do that. However, as that environment is exposed, you can see something better than just the business side of things.

People tend to think, It would be nice if I could see better. Now, this is what people commonly think of as a high-definition image sensor. A high-definition image sensor allows you to see better in fog, blowing sand, snow, etc. than you might think. That's why many people ask why. I believe I can explain it for two reasons. The first reason is the wavelength. It is more of a long wave, so to speak. There is an advantage in that long wave particles can penetrate better. Therefore, when it snows or gets foggy, you can see further ahead. Another important point is that when you drive at night, in reality, the human eye opens and closes its aperture. Most of the time, if the streetlights are shining brightly here, you can see well below, but the sides are not visible. We call it the dynamic range. It's about the dynamic range that we can see.

It's because there is only one gate. What it means to express sanctity is that ADC 14b is not enough, so you have to go up to 18b. But when you go from ADC 18b to 20b, the price becomes extremely expensive. In other words, because there is a limitation in dynamic range due to the presence of only one gate, there is no way to overcome this. However, there is no excellent alternative. There is no external LED, so in the end, there is no alternative. Therefore, you can create an environment where you can achieve the best dynamic range where you need it. So, using terms like foresight and sanctity, you can see further ahead, and because there is no noise level in the surroundings, people prefer using image sensors there. Due to these characteristics, many people have actually tried, and in studios, of course, they are widely used. But what I just mentioned has its limitations.

There is a hidden [MUSIC] over there. Cameras break down every year. They break down. Originally, it's difficult to maintain cameras there for a year. The reason why cameras there have a short lifespan is that they are not silicon-based sensors. There is a significant stability issue. Many products that are not made with silicon-based sensors, so if we make them with silicon-based sensors plus some new materials cheaply and widely distribute them, I thought it would have an impact. I had this thought about 11 years ago. But people are not open to innovative ideas. More than I thought, of course, I worked for a large company before. I worked at Samsung for about 3 years and at Boston Consulting Group for about 1 year. Before starting a business, I experienced corporate life, so naturally, I experienced corporate life, and now I, we, now at Stanford, I am pursuing a Ph.D.

I came and then I think, not that I am the best at studying in the world, but I think I am the best at studying in the world. So, with the foolish idea that maybe I could make the most money in the world once, I started. In fact, if you can't have foolish thoughts, you can never start a business. If you can't have foolish thoughts, you can never innovate. I think that sincerely. I believe that you can only start a business with technology once. If I succeed in this, I will create value by connecting with various companies and businesses in the startup world. Therefore, I think that starting a business from scratch is something that humans do only once. I sincerely believe that. I think that most Silicon Valley entrepreneurs are living like that, and I think that Korean serial entrepreneurs are living similarly. In my opinion, our company's name is the Stratio Project. The project name is the Strato Project. The person's name is Seongwon Strato.

In the upper stratum, I felt a lot like I was being symbolized. When you go up to the upper stratum, doesn't the Earth look like a fist? It looks very small and round. I felt like I had to make a rocket before I die, so I actually started this. I say this as a joke, but I'm serious. So, our company, on the left side, carries mainly what Stratia Korea used to do, and on the right side, it carries mainly what Stratia USA does. In the U.S., we support about 90 companies a year in the form of SB. So, it's amazing that this program has been maintained for about 40 years. SBIR symbolizes Small Business Innovation Research and helps entrepreneurs who have started their businesses based on that. We are proud that, even though we are Koreans, we receive support from the U.S. Air Force, Army, Science Foundation, and the Ministry of Industry.

We received an investment of about 2.5 billion won only in the United States, but it's not a lot in the semiconductor industry. We have struggled a lot, and two years ago, we were the most famous society in the SPI optics. Therefore, we received the Best Sensor Award for two consecutive years at the SPI conference. We decided not to apply this year because we think we should stop receiving awards now. Also, in the United States, there is an accelerator created by people who are known to be cutting-edge in the semiconductor industry. It's called Silicon Catalyst, founded by AMD and Sony entrepreneurs, as well as companies like Fuji Film, and many other scientific companies. We are naturally affiliated with these two devices. If I were to express it in just one photo, this is it. It may look complicated, but I will come back to it in a bit. The rain and the rainbow seem to be different colors, but in fact, they are the same color.

When we move our eyes, it's so cunning. There are white and black people, but they are the same people. It feels like my head is going to explode. Using this feeling of imminent explosion is business, right? All meat is the same. Doesn't the one on the left look rotten, and the one on the right look fresh? It's about creating a situation where it feels like your head is going to explode to make people buy. That's what I think business is. But on the contrary, what strategists do is not to create situations where it feels like your head is going to explode, but to create new objective indicators. That's what I think. When we learn about Jupiter, do you remember the old days? If you think of me as being around the same age, there were no rings around Jupiter in the old days. And on Jupiter, Oh, that's a cloud, what is that? Was there an aurora on Jupiter? And didn't Jupiter have one moon or something like that? It was like that with Jupiter's moons.

There are a lot of them, and there are also belts on Jupiter, and auroras occur on Jupiter for 24 hours. However, only on the equator does it occur. Knowing these facts allows us to understand a lot of characteristics. In fact, the existence of auroras means that tremendous magnetic phenomena are occurring on Jupiter. Jupiter is surrounded by numerous satellites, which are not visible to the naked eye. If we choose the top 15 companies based on market capitalization, among them, nine are semiconductor companies. If they are leading globally, it means they are emerging from Korea. If they emerge from Korea, then large corporations have no choice but to follow. In my opinion, the major trends are AI and semiconductors. So, the money we can raise is around 10 billion won. At first, I am not a chaebol nor the son of one, so what I can do is around 10 billion won. What I can do is to ask you all, what would you like to do with me? I want to create something new with you.

I thought there was nothing else but developing a new application with the device we developed. Actually, it's not just us who think that way. There are many friends doing business with silicon carbide, and there are many friends doing business with gallium nitride. Gallium is for high mobility or sensing devices. That's why we are working on that part. If I can secure a production base where I can create devices with device structures that are not included in the existing device structure tools with these materials, I think I could possibly be in a unique position in the field. So, I started this 11 years ago. Of course, the world doesn't just do ideas.

The world was not easy due to numerous difficulties, but I thought I had to create it eventually. Two years ago, I completed it, and ironically, we have proven it. Therefore, we are transitioning to the lease model stage. In other words, we have completed everything. Now, we need to mass-produce. Consequently, we proposed to four companies in Korea and also to companies in Taiwan. Fortunately, many companies are interested in us, so we are currently making efforts to secure contracts in the most favorable direction. Earlier, the host mentioned about the infrared image sensor. If you think of it as a refrigerator, like an LG refrigerator, you can understand that it has also been applied to apples. There are also semiconductor lines and displays. By looking at them, you can see what defects are there after passing through the wafer.

I think you might also do ES businesses, you could probably do plastic recycling as well as agriculture, right? So, from my side, I can check the quality of the grains, and you might also consider pharmaceuticals. You could also determine whether a drug is authentic or fake in the medical field. Recently, drug detection is moving towards that direction a lot. The drug issue in Korea is very serious right now. I heard that these days, drones are flying drugs from the sea of China to Korea, not just one or two, and if you bring drugs from the airport, they are hidden in candies, canned goods, or even in what was it, vacation items, it's not just one or two problems with drugs. If these drugs cannot be detected, there is no other possibility than using sensors there. So, people who can use that can be useful. That's a small field where startups like us can step in.

Of course, it's small, but it's worth thousands. Not only in large companies, but also in small units, I think the field of sensors is essential. Even in fog, rain, or snow, where you can't see ahead, making cameras like that to be used in cars, drones, defense, and coastal surveillance areas. I believe it's a field that can be applied. I also think it's the biggest in that field, so I'm mobilizing countless people to work in that area. One of the things I hate the most is market reports. Market reports change every year. Anyway, I think the market report is expected to be around 4 trillion won. Since it's not very attractive, I think it will be a business worth about 10 trillion won in the next 3 to 4 years. So, I think [MUSIC] I sold about 8 sensors, so you might be curious about what they are, so if you look, the one on the left is a bit.

Solar panel inspection can overcome situations where the front cannot be seen at the moment of fog and haze. In Korea, surrounded by three seas, it is possible to use coastal surveillance. We need to look at it, right? And it should be attached to the ship. Of course, it goes on the ship. Just naturally. It is essential for vehicles as well, especially in countries like South Korea where the weather changes rapidly. I once went to UA for a short time and drove, and the sandstorm was serious. It was so sandy that you couldn't see ahead. So, I thought it would be useful here too. In situations where climate change is severe like these days, we are presented with more opportunities. I'm not the only one thinking this, right? Many people naturally think so.

Some friends who started a company and sold it to Intel, they finished a three-year vesting period and are now coming out again. When they get acquired, they have to push it for three years. Oh, you know. Yes, the vesting period is over, and now the company they created after throwing away Mobileye is a company called AdaSky. Let me tell you about them once again. What we developed is a far infrared sensor, but they are incredibly expensive, so they don't exist. So, they tried to make it cheaper and created a thermal camera for automotive vision. It's booming these days. As I mentioned earlier, on the left is noise, right? This is all noise, but since there is no noise, this is about the dynamic range you can see in the future. Since there is no noise, you can see everything ahead. Do you understand the meaning of this? Further ahead.

You can see it, and there is no noise, so this is a possible situation. In fact, it shows here. If you don't see this here, it's impossible to see it even in the tunnel. But what is crucial now is that they make it well. So, you might wonder what advantage you have, right? Let's take a look at it again. If you see the same thing, look closely. The lane is not visible here, but when you go outside, the lane is visible. Thermal cameras show temperature differences. The temperature difference between white paint and black paint is what makes the tunnel visible. Since there is no sunlight, the most important thing in a car is the lane. We perceive the distance through the lane. Therefore, thermal cameras cannot be an absolute solution, but even if we have it now, it can be helpful, so we are enjoying it. As far as I know, this company is currently collecting data on autonomous driving cars from six top-tier companies.

So, what we are making can now absorb light of the same wavelength as gas and germanium. Once again, gas and germanium can absorb light of the same wavelength, and gas accounts for 99.9% of the market. The reason is that everything has an answer, right? Making sensors with germanium is difficult, but the process with germanium is highly advanced. With germanium, you can process 12-inch wafers, while with indium, even processing 4-inch wafers is difficult. Because germanium adheres well to silicon substrates, but on indium phosphide substrates, which are rarely seen in the semiconductor industry, you may ask what's the big deal about semiconductors. The moment the substrate changes, you can't do anything. It's really, really absurd. You can understand the meaning if you look at semiconductors a little more interestingly, right?

I heard someone say that there are a lot of spectral sensors other than yours for the second time. We have all done research, and all of them are gas sensors with a minimum price of $7,000. However, we, who have been researching at Stanford for ten years and have been completing products for 11 years, thought we were different. Of course, everyone probably thinks that they are different. I think I am different. But in order to truly prove that I am different with my actions, I have been making sensors, developing products, and selling them. So, in case you find this slightly boring and uninteresting slide, I will tell you something. The biggest reason why we cannot make sensors with germanium is that we cannot make photodiodes with germanium. We cannot make photodiodes with germanium. That's why we...

We have developed a new pixel structure that fundamentally eliminates that. This was our original intention when we were doing our PhD. We have applied for patents worldwide, and we have completed all the global patents. It cost 800 million won for just one patent. Proving that no one else in the world can do this is difficult. That's why many people, recognizing us as rising stars in the new image sensor field, despite everything, seem to be acknowledging us. Other companies have billions of dollars for lasers, investing 500 million won each. Except for us, the ones like us, who can't do it, right? So, I think I'm not good at marketing, but I feel a change in myself starting this year. Because many companies with no skills have dropped out, and people like it when they sincerely fail, putting in 200 billion to 300 billion won.

I understand if I fail when I come out. No one comes out and fails, so I think people are now returning a lot to us. We are classifying about 17 types of fibers, about 91% of them, right now. When you look at this, are you getting the spectrum? The samples were given from the company. Everything there is ours, of course. When you asked if it was possible to distinguish these samples, I showed that it was done well. You might think this sensor is nothing special, but if we don't buy ours, the cheapest one costs $7,000. So, of course, we tried plastic. Plastic is much easier. Looking at colors from white plastic is not easy, but it's much easier with plastic. In the past, we were good at making milk as well.

Oh, yes, even Viagra, we analyze everything. It's a similar principle. When we look at the spectrum from the Swiss side, it's fascinating how we can extract much more information than we expected because of the waves that help us understand the characteristics of the material. It's really amazing for us too. Honestly, when we were scanning plastics, we were very worried about not being able to distinguish them. During the first scan, I made a loud noise, and when the fibers came, I really thought, Is this really different? I thought a lot about it because it was really different. It's amazing. So, why is sign language difficult? Because it's not intuitive, right? I haven't seen it with my own eyes. I haven't seen what color it is. That's why it's difficult. So, we have no choice but to overcome this with our skills. Therefore, we also did a demo for our VIP guests.

If we just play two at the same time, we brought our magnifying glass today, so please take a look later. If you look here, you can't see it when you wear fog, but we can see it until the end. We take pictures directly with our camera. So, I was curious if it really shows when you fog up. If we grow fog, will we still be visible to us? It really showed, so it's incredibly amazing. Even if you look at it, you can't believe it. So, I'd like to say it's like a miracle. Our dream here, of course, is to become the best. So, we are currently collecting government funds in the hundreds of billions unit and also raising money in the VC unit in the hundreds of billions unit. It's interesting that just a week ago, the company came to me.

It's sold, isn't it fun? The market is really fluctuating. We said we would make it like this, but this is a bit boring, so let's move on. As I mentioned earlier, we design, create, and produce it. STN is the name of our company. We've done everything. But now that we've achieved our company's dream, we are in the process of establishing contractual relationships with the companies here to become a lessee. As for pricing, we see the silicon image sensor as our competitor. Our goal is to make silicon image sensors cheaply and have them in every phone. Even Apple is currently putting the Swire sensor on their roadmap. They entered in 14, dropped in 15 because the performance was not good. They are talking about putting it back in someday between 16, 17, and 18. What they need is Face ID, LG's previous technology.

Where do you mainly do it? So, doing face ID in sign language is much more advantageous. It's better in sunlight. Face recognition is done well, but it doesn't work well under sunlight. They say if there is too much noise, errors keep occurring. I want to do that well under sunlight, especially under the display. Removing the reflection on the head is quite an important issue, right? So, I'm trying to set up a camera there. Apple is also putting the installation of sign language cameras in the roadmap for the Vision Pro released this time. It must be because they can capture actions more accurately. Even in a happy situation, so we have some contracts with them. Yes, we have more customers than we think.

We have sold to more than 2,000 customers, and of course, there are no bulk orders yet, so we are struggling. There are still various performance issues, so we are making efforts to create bulk orders. We are currently working on a project with Elsa and a project with the police agency. We have done a lot even without the military. We have done a lot, and what we have here is what we received from the military. The Air Force didn't just give it to anyone, they wanted to use our cameras for something. SpaceX has sold our products, and we are testing them with the U.S. Drug Enforcement Administration. Drug issues in the U.S. are a significant problem, so we are currently attempting a project to detect drugs with our product, which has stronger binding than current legal documents.

안개 껴도, 어두워도 다 보인다! 적외선 이미지 센서의 혁신

안개 껴도, 어두워도 다 보인다! 적외선 이미지 센서의 혁신

스크립트