Radio Cade

Connecting Cars to Traffic Signs

February 10, 2021 Dr. Enes Karaaslan Season 3 Episode 24
Radio Cade
Connecting Cars to Traffic Signs
Show Notes Transcript

Cars that can talk to traffic signs. It’s not science fiction, it’s a company. Dr. Enes Karaaslan is a civil engineering scientist and the co-founder of Connected Wise, a 2020 Cade Prize Finalist. The Orlando start-up is developing technology that connects autonomous vehicles to safety infrastructure, especially in rural areas. Initial data shows that such devices can prevent thousands of accidents per year and save many lives.

Intro:

Inventors and their inventions. Welcome to Radio Cade the podcast from the Cade Museum for Creativity and Invention in Gainesville, Florida, the museum is named after James Robert Cade, who invented Gatorade in 1965. My name is Richard Miles. We'll introduce you to inventors and the things that motivate them, we'll learn about their personal stories, how their inventions work and how their ideas get from the laboratory to the marketplace.

Richard Miles:

Cars that can talk to traffic signs. It's not science fiction, it's a company . Welcome to Radio Cade, I'm your host Richard Miles. Today, I'm pleased to welcome Dr. Enes Karaaslan a civil engineering scientists and the co-founder of Connected Wise an Orlando startup . That is developing technology that connects autonomous vehicles with traffic signs. The company was also a 2020 Cade Prize finalist. Welcome to Radio Cade Enes.

Enes Karaaslan:

Thank you very much, Richard beautiful introduction.

Richard Miles:

So let's start by defining terms. I think probably most of our listeners are pretty familiar with the notion of driverless cars and network vehicles, but just so we're sure. What do you mean when you say connected and autonomous vehicle?

Enes Karaaslan:

Absolutely. We hear a lot about self-driving cars. There are a lot of companies who are actually commercializing marketing very well. The self-driving technology, but more accurate term will be, I guess, for autonomous vehicles. The objective is to give some automated features to the vehicle, to provide safety of the drivers initially. And hopefully in the future, maybe we can replace the driver. There are five levels of automation. What we see currently in the traffic Tesla is one of the pioneer companies and marketing the self-driving. We actually define those vehicles as level two, level three automation. When we reached a level five, we don't need driver on the driver's seat. There are companies who are testing, even doing pilot projects in certain areas in California. In the other States, we have a company who's doing autonomous shuttles in Florida as well. One of the important technologies that will support these vehicles is the connectivity connected vehicle technology, simply aims to provide communication between vehicles and between the vehicle and the traffic infrastructure. They call these technologies V2, I and V2V if it is communicating with the pedestrian, they name it B2X. Now we have a more broader term, which is connected and autonomous vehicle. We would like to provide some connectivity and some autonomous functionality is in the vehicle. So that's how we define it. We are solving a unique problem about these connected and autonomous vehicles.

Richard Miles:

So just so clear on this Enes, an autonomous vehicle that is not connected is something that has, you know, I've , I've noticed recently when I have a rental car, there are these features now where for instance, on the cruise control function, I can set a certain trailing distance, right? When it'll break automatically, as it gets to a certain distance that I can define from the car ahead of me it's features like that. Right? But that doesn't necessarily mean that it's talking to the other cards, it's onboard computer that is performing some of the functions that I would perform, right. And then a connected cars , as you said, more like Tesla, where the car is actually communicating with other cars. In addition to maybe performing some of its functions.

Enes Karaaslan:

I would distinguish vehicle communication from the vehicle automation vehicle automation will give you a features like changing a lane automatically, or stopping at a stop sign or understanding the traffic light color and acting accordingly. The vehicle of communication is a supporting technology that will make a lot of the things easy for autonomous vehicles, such as recognizing a traffic light can be difficult at times. So the traffic light can send a message as real time signal about the status of the traffic light. So before the vehicle arrives to an intersection, it can actually understand what color is going to be when the vehicle reaches that intersection. So the ultimate goal is simply to remove all the traffic control devices that is designed for human drivers that will give us maybe a smooth operation in an intersection. We may not even have to stop in the future. This is the ultimate goal. There are a lot of safety benefits that disconnect vehicle communication can actually provide . They estimate that just intersection crashes, that one single safety application can prevent more than a thousand lives per year. Just one single safety application in an intersection can save a lot of lives. It has enormous benefits for the traffic congestion and the associated costs . So this technology can save us about $170 billion per year, just from the total congestion time.

Richard Miles:

That's a big number. And that was a very important clarification. At least for me, that you made in terms of autonomous versus connected in that these autonomous vehicles are getting more and more sophisticated with their AI so they can recognize a stop sign, or they can recognize a stop light , but they're not actually talking to it . The connected is one level further where the computer on your car is literally talking to the piece of infrastructure and communicating valuable information. It's not just depending on a recognition algorithm, right?

Enes Karaaslan:

We are trying to simply provide as much redundancy as possible to those vehicles. There will be times in a work zone area where there's going to be detour and in this detour, or it will be very difficult to navigate safely because the lane lines disappear. Sometimes maybe a flagger is simply rotting the traffic. So in complicated scenarios, the vehicular communication can be very helpful. We can send real-time signals, wireless signals to those vehicles about what they are supposed to do and how they should act in a traffic situation like this.

Richard Miles:

So Enes let's talk specifics of your company and what you're developing. And I think that would be useful for listeners to understand that a lot of this depends on infrastructure investments by a given city or municipality. And from what I understand, there's a big difference. Now there's a big gap between the smart infrastructure that you've seen urban areas versus rural areas. So could you give us sort of an idea how big is that gap? And then what sort of timeline are we looking at to catch up a small town in Ohio versus large city like LA or Washington DC in terms of the infrastructure to support smart technologies or connected technologies.

Enes Karaaslan:

This is a very good question. This is the actual problem we are hoping to solve this vehicle communication technology that I just explained uses wireless signals, and it requires fiber optic infrastructure. And a lot of the times, especially in the rural areas, we don't have any of those. So bringing this technology to the areas, we actually will need the most because rural areas experienced higher traffic fatalities than the urban, our countries roads are actually 97% in the rural. So we have large rural areas that we want to bring safety. The main problem is going to be very expensive and not practical to deploy this technology in these areas. So we approached the U.S. Department of Transportation with this idea of using traffic science. We told them we could use simply these science , smart science to allow communication from the infrastructure to the vehicle. And we can help support the existing communication and as well as we can help autonomous vehicles in traveling in the rural later, the U.S. DOT, liked this idea very much. And they awarded us Small Business Innovation Research award , and we've been putting effort for the last three years, doing a lot of outside testing in the challenging climates , challenging conditions, sometimes different speeds. So we reached to a point that we can reliably use a smart sign like ours. They look like colorful cue QR codes, but they are designed specifically for this purpose. It doesn't require any wireless communication. It simply sends a message and information that we need to send through the vision-based communication. There is a camera inside the vehicle. It recognizes the sign and decodes this message.

Richard Miles:

Let me see if I understand this correctly. I saw a demo video or one of your devices, and this looks like something like, like a very large transponder that people are used to putting in their car for like ETolls and stuff like that. Does it visually recognize, say a code on a, say a stop sign or any other traffic sign. And is that how it works or is there an active signal that is being sent out?

Enes Karaaslan:

It's a visual identify a visual code that is linking to a message. When this visual code is identified as an encrypted code, only the device can understand what it means. So a third party cannot alter the message, but we decided to make also a device that is not just giving this capability, communicating between the sign and the vehicle, but also give the driver some of the advanced driver assistance features. So when you put this device on your windshield, it's not just going to recognize those signs, but it will also give you a lot of the features you will see in Tesla. It will recognize the traffic signs and act accordingly. If your vehicle support automation, you can optionally use those automation features too . So your vehicle will stop when there is a red light detected. So we give all those features on top of we provide this vehicle of communication between the smart sign and the device. So U.S. DOT need was the actually, how can we move this connected vehicle technology to rural areas, but we also needed to consider some commercial aspects of it. How can we market this device to the current drivers? So we thought if this device could also give some advanced driver features, the drivers will definitely benefit from it. We are aware that it's going to take some years until we deploy these signs to the locations where the DOT is going to need until that time our customers will be able to use those devices to benefit from some of the emerging technology features.

Richard Miles:

I'm guessing though, that the device requires some sort of software interface for the vehicle itself, right? So is that relatively easy to do? Is it something like an Apple play feature where you just plug it in or communicates wirelessly and then it takes the information that it's getting from say a sign. And what does it do with that information? Would the car, for instance, automatically slow down as it's approaching a stop sign? Or is that how it works?

Enes Karaaslan:

Yes. If your vehicle supports some automation, usually 2016 and about models have that capability, then it can simply send the vehicles control system, a signal, a message about a traffic situation. But if your vehicle doesn't support any of it, it will simply warn the driver. We are the visual audio warning. So it's going to tell you and navigate you on this complicated work zone. I'm not sure if you're familiar with Florida, but we have a big construction project going on I-4. And when I try to use my Google maps, it always fails. The road is changing constantly. So if we could put one of these signs are for the, these devices will simply relay the correct information about the road construction and the navigation system in that screen, you will see will guide you on the accurate route. We also made a lot of useful applications, scenarios, not just in the rural areas, but also in the urban areas. We were demonstrating in this automated vehicles summit in Florida, it was in Miami downtime . One of the applications was a garage parking application for handicap people, transportation challenge people. We simply put up one of our signs in front of this garage, very big garage. And it was even very difficult for us to find a handicap parking. So as soon as the sign is recognized, the device simply navigates the driver to the closest handicap parking area. We do a similar application for electric vehicles. Sometimes it's difficult to find a charging station. So it navigates the driver to the closest charging station in the garage. One of the good application for urban areas was a lot of the times the navigation apps fail to give you the accurate route because those high rise buildings, block the GPS signal. However, the smart signs can actually be helpful in machine vision mapping. What it is doing is when the sign is recognized, it can navigate you on the accurate rod, even though you don't have any GPS signal.

Richard Miles:

I see. And one of the things I like about doing this podcast is I think I understand the technology. Then I talk to the inventor and then I really understand it cause I didn't focus or I didn't get the fact that this differs from say, Google maps or GPS function in that one, depending on GPS, obviously. But two , those are maps that might be out of date the next day after they're uploaded or you get your new software. Whereas the Connected Wise devices, depending on either recognition or current stimuli, I mean it's information, it's their current at the moment, right? It's never out of date really, right? Because it's reading what's in front of it in a sense.

Enes Karaaslan:

Yes, there is a dynamic condition like work zones. And if there's a road closure, simply the construction company puts up one of our smart signs and relaying the updated road chemistry message to these devices. So then your vehicle sees the sign. It will learn the accurate route and it can navigate safely on the detour. And it can understand the situation about that road closure. There was even one interesting application with demonstrated for the urban area to Google maps was actually routing the traffic and telling the driver to make a left in this intersection, but there was a no left turn sign there. So the route was an updated of course, about this traffic situation. What we did was stood up one of our signs and we rather the traffic on the accurate route by giving the updated roads, government tree and the traffic condition in that area. So we are trying to build something that traffic operators can actually help these connected automated vehicles and let them navigate safely in the challenging traffic conditions or in dynamic scenarios like work zones. It's building communication between the traffic infrastructure and the drivers.

Richard Miles:

I see. So really the most advantageous in situations that you have to depend on real-time information, because there's been some change like a work zone, as you said, or for some reason the GPS driven database might be out of date or there's no other infrastructure to support smart technology.

Enes Karaaslan:

And also this message we are sending is not just a smart sign is not simply saying what type of sign is going to be. It's saying a lot more than a normal traffic sign can say. So it's not saying just no left turn. It's actually giving the whole road geometric data of that intersection, which is very valuable for autonomous vehicle because intersection situation, the current sensors on the autonomous vehicle can be challenged sometimes to understand the road geometric history is happening there. A LIDAR may not see the other side of the road in a lot of situations. So if he could send the road geometric data, to autonomous vehicle via a visual input, that will be very helpful for those vehicles and help them localize their position in terms of a non road geometry data, right? So we are doing different products, different market for different market segments. We have different products feel about autonomous vehicle. We are simply aiming to sell a software solution since your vehicle will be accurate with all those sensors we need. So your camera will be available, et cetera. It will be as simple as a software solution. Then the vehicles subscribes to that software, it can start recognizing those signs. If you have a conventional way called that is not a be any sensor, any camera, then we simply sell these devices. They are very affordable. They cost under 500 bucks and it gives you a lot of automated vehicle features. It uses state of the art, artificial intelligence models. What we achieved is really great. We were able to put all those complicated competitions in a very small affordable device. And so that's what we are hoping to achieve in this project we have in the company.

Richard Miles:

So Enes, as I always like to hear about the kind of the aha moment from inventors or founders. So tell me a little bit about how, when and where did you, and I don't know if you have co-founders, but did you come up with this idea? Was it just sort of a flash of inspiration or did it slowly dawn on you as a iterative process?

Enes Karaaslan:

So the idea of using signs to send a message has been around for a decade. The challenge was they tried matrixx barcode kind of system, similar to QR codes. There were big companies who attempted to use these kinds of technologies. The problem with the QR code base technology was sometimes they are designed for laser scanners. And if you want to use a camera, it's not going to be working as robustly as we hope , even with ourselves is very common QR code applications in our mobile phones. But when it comes to a traffic environment, there are so many challenges, dark time, nighttime condition , bright sunlight. A lot of the times only a small portion of the sign is visible to the camera. But in the QR code based systems, when you cover a small portion of the message of the barcode, the whole message is running . We put a lot of thoughts to solve this challenge. We need to use the power of image recognition because this barcode systems are simply encoding decoding methodology. If you simply trying to understand the black and white areas, and if the sign is partially covered or not visible, then it won't happen. This is a real common application in the internet forums. You will see when you sign up in a forum website, you will see a default avatar in terms of shapes. They call it identity cons . They are automated . They generated images unique to your IP address. So that was our aha moment. Okay. This is a unique image generated automatically. And it's this thing to a V IP address every user on the internet. So we decided to generate unique images for sign messages. Every time we put up a sign, it's very distinct and the other sign we built , they are not a barcode system, but they are actually unique images that is simply a visual identifier for a message. So this gave us so much capability in terms of even 90% of the sign is not visible. It can still safely distinguish the sign from the other signs and activate the correct message. Now, we were able to use conveyor affordable camera system to operate safely and high vehicle speeds, or even in challenging elimination condition. When the sign is not very visible, even when a tree branch is covering the sign or snow is covering the sign, it can still pick up the message robot slate . So that was our a home moment . I was, I guess, cruising on a forum website and realized that, okay, this is an automatically generated that I could actually use to solve this problem. And U.S. DOT really liked this idea. It was very different, very unique than other ideas. That main focus on to our QR code, bar code kind of systems.

Richard Miles:

Right? That's fascinating, but I'm sure you've discovered that great ideas don't sell themselves. So tell me a bit about where you are in terms of a company you form Connected Wise. Tell us when you did that. And then where are you in terms of the development of your technology in terms of funding or employees or path to market, what does that look like for you?

Enes Karaaslan:

We started the company in 2018. When we received the grant from the Department of Transportation, we were a very small team of three to five people all founding members for PhD graduates from my University in Orlando, the University of Central Florida. We did an extra ordinary performance. We did a lot of testing outside, we're using our own vehicles. So we did a prototype that was ready for demonstration. After phase one U.S. DOT really like our performance decided to award a phase two award, which was a major grant for us. We of course grew our team. And now we have 12 people in the team and two big offices trying to commercialize the technology. Now we made a lot of progress in the technology side, but making the technology ready for a commercial product takes a lot of time. You have to think of a lot of marketing commercialization. You have to think of advertising. And our go to market strategy took some time for us to figure out who our initial customers are going to be. The majority consumer. We decided to first target the city and counties started from Florida's rural counties and our simply approach to that. Um , for pilot projects, we asked if you could put up these signs and deploy some of our devices in the vehicles and do a pilot project and see how much safety benefit we can bring to the county vehicles. And in the continuation of this pilot project, we can distribute some of these devices to the volunteer residents of the County and measure a broader scale benefits of the technology. So we have done a lot of communications with Florida county here, they are transportation departments. We are hoping to start a pilot project in Florida and very soon. So those are our initial customers. However, in the future , we are going to be targeting fleet customers who have a fleet of vehicles that we can simply provide advanced driver assistance futures . By that time, we are hopefully going to build some science in several locations in Florida. And these fleets also can benefit from this vehicle communication futures as well. But the main goal will be for fleet customers, helping them to collect data from their roadway . Sometimes for road infrastructure operators, our devices can collect a lot of data from the roadway about the work zones or the traffic congestions , et cetera, even the asphalt condition in the late maturity , we are hoping to target automakers who will agree to integrate the technology natively in the advanced driver assistance technology, we'll be able to support our system and it can recognize our signs in the future. Hopefully this is a multi-faceted market. So in the other side of the business, we are to license the sign technology to the sign manufacturers because they are also interested in connected vehicle applications. And there are billions of signs in the world and the placing those signs is a huge market for the sign manufacturers. So that's the other phase of the market. We are hoping to target. We are aiming to make $50 million by the year 2025 selling around 50,000 devices. So that's our objective in the near future.

Richard Miles:

So it's a very interesting point. You mentioned earlier about data and data collection. I didn't understand how valuable that is until my daughter who an actuary for a major insurance company explained to me that getting really high quality data, especially in automobiles is of enormous value in particular to insurance companies who are very, very interested in all of those details. So I didn't realize just to have valuable that is, and it's in terms of investment. Now you mentioned the Department of Transportation a nd that grant, are you also raising money from private investors? Or how does that work?

Enes Karaaslan:

So we are now seeking seed stage funding that will be in the form of matching contribution. U.S. DOT Is hopefully going to award another round of funding for commercialization efforts solely, and its seed stage funding will be simply matching that contribution from the government. It's going to be a safe investment for the investors saying that you're only going to invest if the government is interested in the technology and decides to invest for commercialization right then by the end of 2022, we are hoping to raise series a around funding.

Richard Miles:

So that's a pretty rapid timeline. It sounds like you're well on your way. And it's one final question. I always like to hear about the personal background of inventors and entrepreneurs. Tell us, what were you like as a kid? Were you one of those kids on the playground that made the rest of us feel stupid because you were building suspension, bridges out of twigs and the rest of us were just playing with our toys, or what point did you want to be an engineer or what was the path to your current career when you were a kid?

Enes Karaaslan:

So I always known that I would be an engineer even when I was a child. So I guess that's something I was very sure about. I was always creative building things was always hands-on, but my major is civil engineers , but I always wanted to be a computer engineers . However, if you're an engineer, you do a lot of the things. You can be a great computer scientist. It doesn't matter which engineering discipline you are in. We are all using the same tools, same knowledge in our fields. What I was good at, how I both, I was good at knowing a lot of things, rather than being best at one engineering discipline. My uniqueness was I was able to connect different things with each other. Computer science expect with a civil engineering practice. And even in this case, it was an IP address future to a transportation application. So that was something I was good at. It's sometimes a PhD students. They usually have this hardship and explaining complicated things to the public. And I believe I was better at compared to my peers explaining the complicated concepts in a simpler rehearse to my colleagues or to my friends. So I have been always an entrepreneurial person than fans in my class. I started company when I was still a student. So I finished got my doctorate degree, but my talent I guess, was to be able to connect different things that are in different disciplines.

Richard Miles:

Enes, telling me, does this run in the family were either of your parents, engineers, or business people or any of your siblings?

Enes Karaaslan:

My father is a great engineer. I guess I got it from him because he was an electronics engineer, but people expect them to do even a good job in changing electrical system in household, but it's not his expertise. But what he was good at is even though something is not his expertise, he still think that is his responsibility to be good at things that public will expect from it. So he was able to fix any electronic equipment in the house. He was able to build his own furniture. He was able to do a lot of the things by himself. And that's something that I admired , I guess I wanted to be in the same way. I was a civil engineer, but I never said a climbing job is not my expertise, even though public think that it could be associated to my field.

Richard Miles:

Right.

Enes Karaaslan:

But I try to learn as much as possible from different things. And I guess what is unique is sometimes is if you connect in a different field, something you're good at to another field that you're also good at, you can achieve a really great thing by communicating these two, right? And a lot of the emissions in science happen in that way, when you get different disciplines together.

Richard Miles:

Right? That's something we'd talk about all the time at the Cade Museum is how invention is really making connections between a fact over in this field with a factor in that field or an insight, and a lot of different inventors have a lot of different interests and they connect a field that they know to another one. And one final question. I usually ask this earlier, where did you grow up?

Enes Karaaslan:

I grew up in Turkey. I moved to United States to study my PhD six years ago. I had been here before, during my undergrad education as an exchange student. And I guess during this exchange semester, I made a lot of good friends who later visited me back in Turkey. So that was, I guess, helpful to my decision of coming back here to continue my education. I also admire the competition here in the higher education, millions of people from so many different countries competing something greater. So that really attracted me. And I really enjoyed that competition here .

Richard Miles:

Well, I have to say you probably couldn't have picked a better city than Orlando. I'm not from Orlando, but I've been there quite a bit. And in terms of cities there's growing that is developing, that is trying out new things. Orlando certainly has a lot going in that direction.

Enes Karaaslan:

For entrepreneurs is growing really fast, especially Florida is trying to become an autonomous vehicle hub.

Richard Miles:

You couldn't be in a better location and it's thank you very much for joining us today and Radio Cade, fascinating discussion and your company's doing well. You're doing well and wish you the best of luck.

Enes Karaaslan:

Thank you very much. I really enjoyed this podcast.

Richard Miles:

Great. Thanks for coming on.

Outro:

Radio Cade is produced by the Cade Museum for Creativity and Invention located in Gainesville Florida. Richard Miles is the podcast host and Ellie Thom coordinates inventor interviews, podcasts are recorded at Heartwood Soundstage and edited and mixed by Bob McPeak . The Radio Cade theme song was produced and performed by Tracy Collins and features violinist Jacob Lawson.