Episode 211 - Is Radar the Most Underutilized Software?

Flexible, adaptable and upgradable. Not to mention cost effective and scalable. With all of this potential, the full value of radar has yet to be unlocked—until now.

Built on a proven military grade software, Spartan Radar is harnessing the power of mobility sensing for the safer and more reliable movement of people and things. With the highest clarity in the radar image, Spartan’s proprietary software works with existing sensor stacks to enhance resolution without more expensive hardware. The powerful combination of radar, cameras and software can bring cost-effective safety and autonomy to the masses.

To learn more, we sat down with Dr. Matt Markel, CEO, Spartan Radar, to discuss why radar has yet to be fully utilized in the market, the amazing potential it holds in democratizing safety, and how camera and radar can work together to unlock the full power of mobility sensing.

Meet Our Guest

Spartan Radar

Dr. Matt Markel has excelled in engineering, technology, and leadership in both the defense and commercial sectors. He created and led the radar division of Ghost Autonomy, an autonomous driving company using breakthroughs in collision avoidance technology to develop safe, attention-free self-driving for mass market consumer cars. Prior to Ghost, he led the radar team at Waymo, formerly the Google Self-Driving Car Project.

Dr. Markel is widely recognized as an expert and leader in radar, especially automotive radar for autonomous systems, Technology Readiness Assessments and Technology Readiness Levels, and electronic warfare. He is the inventor on multiple patents and trade secrets from his time at Raytheon and Waymo, with patents pending from his work at Ghost Autonomy.

He is a published author of Radar for Fully Autonomous Driving, and is now the CEO of Spartan Radar, a radar software and hardware startup based in Southern California.

SAE offers a full portfolio of information products based on topics just like the one you listened to in this episode.
Contact us to learn how we can help advance your business.

Contact Us


Grayson Brulte:

Hello, I'm your host, Grayson Brulte. Welcome to another episode of SAE Tomorrow Today, a show about emerging technology and trends in mobility with leaders and innovators to make it all happen. On today's episode, we're absolutely honored to be joined by Dr. Matt Markel, CEO, Spartan Radar. 

On today's episode, we'll discuss Reese's Peanut Butter Cups and how they equate to camera and radar. That's right. The camera and radar are the Reese's Peanut Butter Cups of autonomy. We'll also discuss the cost efficiency of radar, the applications for radar from commercial vehicle to ag to SAE level four autonomous vehicles. We hope you enjoy this episode. Matt, welcome to the podcast. 

Matt Markel:

Hey, Grayson, thank you so much. It's a pleasure to be here.

Grayson Brulte:

It's great to have you here, Matt. You're a world-renowned expert on radar, all things radar. And I'm really curious, there's a lot of discussion when we saw each other at CES around radar, the improvements of radar, all the breakthroughs that are happening there.

Matt, in your opinion, what is the current state of the radar market, and what efficiencies is Spartan unlocking in that market? 

Matt Markel:

I think it's a really cool time for radar. But it isn't what everyone thought it was going to be, say, five years ago. On the one hand, Radar is prolific on passenger cars for L2 features.

There's lots of research and development going on as well from different players. Obviously, the players in the L4 market, the Waymos and the Cruzes. But you're also seeing Radar startups, obviously Spartan Radar, like my company, but others as well, like Zendar and Perciv, as well doing some really cool things.

And the silicon makers like Arbe and Under are continuing to push the envelope as well. We're also seeing some work from the conventional tier ones, and obviously Mobileye is still doing well there. With the commercial vehicle market, we see Radar sometimes being used across different vehicle types and classes, although the technology there is still pretty limited.

So really good things from the proliferation standpoint, but it's also a little bit of a weird time. We're trying to figure out what the next big pull is for radar. Obviously on commercial vehicles, getting those up to the level of passenger cars that we have today. That's a big thing. And then their environments are so complex, seeing how they're really going to lean in, into these different environments and these different, off road, on road applications as well.

The L2 and L3 needs are still increasing. We're seeing increases in NCAP capabilities and so forth. But you're not seeing the imaging radars everywhere across the solicitations like we thought there was going to be five years ago. And that's the big changes. Not everything is going to be this fire breathing imaging radar.

And of course, there's a lot of external environments as well going on. Strikes, inflation, etc. It's a really interesting time. As far as what Spartan's doing about that, our belief and our strong belief is like software being software defined is the key enabler there. And there's three things that I think are canonical takeaways from being software defined.

One is flexibility, making the same hardware support multiple functions. There's adaptability, which is having the way to optimize. For the current situation that it's in. So adapting in real time to the environment that it's in, and then upgradability, which is adding new features to be incorporated over time, which extends both the useful life of that particular radar and the useful life of the product, but also the useful life of the production line as well.

Grayson Brulte:

That's really interesting. So I like that flexibility, adaptability and upgradable. So let's just take a passenger car. For example, are there different types of radars on a passenger cars? Or different type of software. What does that look like? 

Matt Markel:

So right now there are usually multiple different types of radars. What we see in the front is usually more of a long range because of the speeds of vehicles are going on. That needs to be a longer-range sensor. But what we don't see, what we don't need is that same range capability all around the vehicle. Sometimes we call those these surround radars. We don't see that long range capability there.

Grayson Brulte:

Fascinating, because you ran, or you, I think I could properly say you built the radar department at Waymo. And that, it was really interesting, so I was doing some research prepping for this, and your robo taxis that you helped develop had six radars surrounding all sides of the vehicle. What kind of performance did that unlock?

Matt Markel:

Great question. So Waymo had a really advanced sensing suite. We had cameras, we had lasers, we had radars, we had microphones, on the radar side, the everything, all the data from the radars were processed centrally. Which is really great because it allows you to have this universal perception view.

And that to me, I think is the, one of the really cool things, all the sensors come in together and you can see that from this one, holistic view of the world, of the scene. In addition, with some of the things that we did with the, with those radars, where we had overlapping field of views in critical areas, which from the radar standpoint, that increases the update rate.

That's also increases the probability of detection in those regions. So that by placing the sensors appropriately, you can get basically better coverage and better performance in certain key critical areas. The other things that the sensors at Waymo were like really good, fantastic sensitivity. I don't know of another radar out there.

That has the same sensitivity as the radar that we built at Waymo. It was the, it was really fantastic and great spatial resolution as well as purity in the scene from the dynamic range standpoint across the board. And I think everything with Waymo was always centered around safety and the safety mindset was super strong.

So when we provided information, we provided not only the, did we see something or not, but everything was. done with the concept of, should you have seen it? If you, it's one thing to say, ah, I didn't see anything, proceed on ahead. But if you wouldn't expect to see something because you're blinded or there's something very large in the area then that causes you to have to make a different decision because now you don't know, right?

So if you think of like, how you're driving and there are trees off to the left and you can't see what, you can't see anybody coming. But you know that you can't see anybody coming because there are the trees there. So we brought that same mindset to the Raiders as well, but it was just a, a really awesome sensing suite that we built because it was a robo taxi application.

Grayson Brulte:

How do you take that same performance, same focus on safety and put it into passenger vehicles? Obviously, your traditional passenger vehicle Waymo like budget for all the. It's called a sensor sweep, but how can you take that performance? This is a software breakthrough on the radar or what? How do you do that?

Matt Markel:

So remember that Waymo, the business model is that of a robo taxi, right? But for passenger cars, we're expecting someone to write a check or at least sign up for payments on day one of ownership. So it's different than a robo taxi and cars are expensive. Oh my gosh, cars are expensive now. And I saw an interesting statistic last, recently, that like in Q4 of last year, one out of 13 auto loans moved into delinquency.

So that cost of these auto, automobiles is catching up with us. So we need to keep it affordable. Otherwise, we won't be able to make, if we're concerned about safety and taking a bite out of those 43, 000 lives that are lost in the roads every year in the United States, we need to put this technology In vehicles that people can actually afford, right?

If it's, if everything's only in 150, 000 car we're not going to make a big dent in the penetration with that penetration in the market with that. So we need to put this in things that people are going to be able to afford. So what we found out and Spartan really worked hard on, on cracking that on kind of these use cases is that with a lot of the ODDs.

You don't need that super long-range capability all around the vehicle. You need it in the front, but you need it all around, but you need imaging all around the vehicle, but not the super long range. So what does that enable? So if you want this imaging capability, You could use lasers, but you can't really because they're very expensive still, and you could, we can't put the imaging radars there because those are still at a higher price point.

So you can use these small corner radars, these surround radars, but you need to increase the spatial resolution of those. And that's one of the things that we are doing at Spartan is. Of all the things that radar is good at, spatial resolution is the absolute worst. We have a software layer called Clarify that helps to beef that up.

Basically, increase that spatial resolution for the system. Like eyeglasses for radar. And, with that, coupled with the cheaper radar, then that becomes something that we can start bringing this to, to passenger cars. 

Grayson Brulte:

Okay my wife's gonna choke me for this, but I gotta go there. Eyeglasses for radar. She calls them her cheaters. She has to get out her cheaters to see something. Then she goes to the Publix and gets her cheaters. In a hypothetical world, can you put cheaters software on the preexisting radars that are there on cars to allow that better vision? 

Matt Markel:

So it takes a little bit, it's not exactly as easy as going to Publix or CVS and buying a pair of foster grants or something like that.

That I might have done at one point. I actually decided that once my eyes started to get to the point where I needed to wear glasses all the time, that I would go all in on like this. Fancy eyeglasses and stuff like that. So I bypassed the Publix and the CVS is now and spend far too much money on eyeglasses, but it takes a little bit more integration than that but in general, our layer, our software layer inserts somewhere between the range Doppler compression and the final beamforming layer.

And so it's a little bit of, it's a little bit of, it takes a little bit of engineering to get in there, but we've gotten over the last few years to the point where we can optimize that for a particular radars and tenant topology and compute topology really in, weeks to small numbers of months. So in terms of the lifetime of building out a product, that's not the limb fact. 

Grayson Brulte:

No, no timeline on this, but in a hypothetical scenario, you're working with a tier one supplier, you're working with an OEM and Spartan, let's say Spartan has a software breakthrough, can you push that as a over the air update to a vehicle that has a certain type of radars on them in the future?

Matt Markel:

In the future, you should be able to, that's correct. And that's that upgradability of the software defined radar concept anyway. It's like we, as you develop new features, as you develop new capabilities, that you can then roll those out, to the cars that are out there today, which again, extends the life of that car, extends the life of that how long someone's want to keep the car.

So that's the idea is I eventually be able to be done through a, through an OTA update. 

Grayson Brulte:

That, if you look at it from a, you mentioned the auto loans which is, it's scary. It comes down to cost. So this, it seems like you can increase performance and increase safety in a more economical fashion. Is that fair statement? 

Matt Markel:

That's exactly it, right? We've got to get the safety to be affordable. We've got to be able to bring this to, and by the way, we talk about safety all the time. And I think that's really important, but we're also talking about. Quality of life, quality of drive features to whether that's parking assist, whether that's the, getting into the higher levels of autonomy and passenger cars.

Those are important as well as just basically staying alive. Yeah, so the idea is that we want to bring that to as wide of a market as possible and to do that. It has to be cheap and almost anything that's cheap has to be done through software as opposed to expensive exotic hardware solutions.

Grayson Brulte:

Is software the key to unlocking the true potential of radar? 

Matt Markel:

I would say it's the new thing, but it's not new. I remember the first software reprogrammable. So as I came from the defense industry and In, I believe it was 91 we IOC in initial operation operating capability. The first software reprogrammable Air-to-Air Missile, the Amram the M1 20 B, and which was a game changer.

Now you can get this, the hardware out there, which is the long pull on things, and you can put a different software load in that, the initial one that you wanna get out there. And you can have a planned, we often call those pre planned product improvements or p3i efforts. So you get the hardware out there.

And you've already laid in this road map for what you want to be adding in overtime and of course you can change that concept you can tweak things along the way we would lay in things that we knew that we wanted but also have the flexibility if the bad guys did something that we could adapt to that and had that all there and that's all enabled through being software reprogrammable. 

Grayson Brulte:

Was software reprogrammable or software defined radar invented in the DOD sector?

Matt Markel:

Great question. I, from my experience, I think it was, I think that the, because DOD was such a, just such a dominant force in radar development from, for several decades, 

Grayson Brulte:

I'm going to stay on defense here. It's really interesting. You have a defense background and so do a lot of the employees that work for you at Spartan have defense. Why? Is it the skill set they bring, or why are you hiring so many individuals with defense backgrounds? 

Matt Markel:

Great question. I really loved my time in defense. As a matter of fact, so I started right out of undergrad in nineteen hundred and ninety in working on, missiles and really cool radar related missile things.

And I didn't think I'd ever leave defense. I thought I'd be doing that. I pivoted throughout my career into different roles, but I never thought I'd leave the defense industry. And, it's an interesting story about how I, what caused me to leave but I think a lot of folks at Spartan thought that way too, and the skill sets that you develop in defense are really interesting.

And I think that there's a little bit of a difference between those and what you would see in just in straight up tech. One is we bring a deep system engineering focus. And we think holistically about the system, not just as a widget or an algorithm, or even a sensor, think of the problem from start to finish, and this is a, this is super important, find a way from start to finish to see at least one possible solution.

So you can show that the problem is tractable. That's one of the big things that we would do in defense. The other thing is. Another thing is like when you think about cutting edge technology, we were, we would use cutting edge technologies to solve a lot of problems but we also knew that the cutting edge technologies are, although they're a fantastic tool, they're just one tool.

They're not the only tool. So sometimes simplicity. Is more important than the fancy technology. And then this is back into the systems engineering approach. So I worked on some systems that were these radars that were, that go on ships. They were like 14 feet tall and had banks of, yes, big systems. They had these big banks of computers all doing the processing and on the signals that were coming back and figuring out what to do next with the radar, because it was very agile and what, how it could approach things. And clearly, really advanced technologies in that. And that's, and you need that because the, the, to support the ships and support the fire control and those as well.

I also worked on some things that were, had to be really small and very efficient with power usage, because, but they weren't cheap. They were they were less expensive, but they weren't like this doesn't this can work, or this could not work. They were super critical to certain missions.

Thinking through the problem and developing the right solution is a super important skill. I guess the other thing is we think about the strong leadership roles we typically had in defense. There were, a lot of times we, people who'd served in the military were leaders in the industry side are definitely in the streets leaders in the government side as well.

So you get strong leadership role models in this case. It's concept of purpose and concept of serving something bigger than yourself. 

Grayson Brulte:

It's for the good of the country, it's for the good of society. The DOD plays a very critical role in keeping, keeping us safe. The unclassified answer, I know the, I'm not going to ask for the classified answer, I know the lines.

The unclassified answer, what is the coolest thing in radar that you worked on that defense that is unclassified? 

Matt Markel:

From an unclassified standpoint of what, of the answer I can give, the, one of the coolest things. that I worked on was in some really advanced stuff on missiles and missile radar because it's a Because it's got everything right?

It's got the Not only do you have to have a radar that is in something that ultimately is going to be You know, in a matter of minutes, not there anymore, so you have to be understood that cost point, but also that missile is often protecting something that's worth hundreds, thousands, million times more than the missile itself.

So you have this weird cost parity that you think you need to think through. The other thing is like everything that goes into the missile, you've got the size, obviously with the, with radars, the bigger the aperture, the narrower the beam. But you put your constraint on the size because this might need to be inside the bay of an F 22 or an F 35 or something like that.

So you can't just make it arbitrarily large. And then just the concept of and this is going to be going, supersonic or perhaps even hypersonic. And just the, how that impacts things like the uneven heating on the radome or having to see through different parts of the radome throughout the mission and the flight profile and ultimately tying all that back into the guidance algorithms that allow this missile to ultimately impact the target.

It's, there's so much cool technology stuff all there together that it really makes for interesting. Interesting engineering and an interesting systems engineering problem. 

Grayson Brulte:

Because you mentioned the background of your colleagues that they're, that deep system integration, I'm thinking, okay, g force, you have wind, you have weight, you have heat, you've got bad guys trying to knock it down, you've got all sorts of things happening.

Matt Markel:

Jammers, people trying to jam it. Yeah, exactly.

Grayson Brulte:

So you have to engineer from a systems engineer start, you would have to I'm going to use a Silicon Valley, sit up there with a whiteboard, think of every possible thing that could go wrong and then engineer a solution for that when you were developing this? 

Matt Markel:

You would start with like kind of what some of the constraints were, right? So let's say we're going to make a new version of an AMRAM. You might be constrained that it has to be the OML, the outer mode line can't be any different. So that's one of the constraints. Now, how much do you want to have dedicated to the motor? Versus the warhead versus the guidance and control section, the electronics where the radars and stuff are up front because there was this great PowerPoint slide that I've got somewhere in my files that says missile design from the viewpoint of these different engineers.

Like the, from the aerodynamicist viewpoint, it was a, it was like a dart, it was from a drag standpoint, it was a dart because it was like had the least amount of drag from an aerodynamic standpoint of controllability. It had these giant wings, so it gave it the most control surfaces.
So you have all these different factors in play from the seeker standpoint, the radar in the front, it had almost no motor, almost no warhead, but this giant radar in the front of it. So all these different concepts were there and everybody has to find it out to figure out what's the optimal blend.

And of course, that's driven by what are you actually trying to accomplish? Do you, are you trying to, what's the type of targets you're trying to hit? How far away will we be launching them? If you launch at this range, is the shooting aircraft in. Is that in jeopardy or is that, able to stand off still? So there's a lot of different aspects that go into the systems engineering trade. 

Grayson Brulte:

It's fascinating. I read the book Chip War recently about the Texas Instruments, commonly known as TI, the amount of chip engineering that we couldn't, I think it was in Vietnam. We couldn't hit a bridge and they put a TI chip in there.

Next thing there's precision accuracy. That's, it's really incredible. This engineering that goes on inside of defense and eventually then goes the commercial sector. You have this background in defense, you got to play with really cool toys and solve really complex problems. Then you go to Waymo, you build the radar division at Waymo, now you're running Spartan.

How has your background in defense influenced your thinking on radar? Are you taking that same mindset across the board on all your journeys? 

Matt Markel:

There's a quote by Wayne Dyer that says, when we change the way we look at things, the things we look at change, which is really interesting when you think about it.

And I've been blessed in my career to be able to see things from multiple different vantage points, from working on the government side of the coin in defense when I was at the Air Force Research Lab and their special projects group to the many sides and facets of Raytheon when I was there for seven years working on big programs.

And then eventually becoming the tech director for the advanced electronic warfare P& L group, and then working for the CTO. There's lots of different things there to Waymo, which was a medium sized startup and then a very large startup. You're going for 500ish to 2000 people by the time I left.

And Ghost, which was a smaller startup. And now to Spartan, which is an even smaller startup, you get to see all these different technologies, but also all these different mindsets and understanding how the mindset is and how the, how people are viewing the problem. For example, are they, what are they optimizing for? Is it risk aversion? Is we need to get this out there, performance speed, who cares as much about the performance, but it has to hit this, it can't go over schedule, or is it pushing the technology, this whole cost schedule performance triangle is understanding what's the most important on that is really is hugely enabling.

Ultimately, how are you going to use that to meet the needs of your customer and whatever you're trying to push for society as well. How does that influence my thinking? I think, on the obviously I'm very bullish on radar still the shocker, it's, I'm still very bullish on radar.

But I think If we extend that up to autonomy, I think it's really interesting and I think that it really has to be a yes and approach. The more I've been thinking about autonomy, it has to be a yes. And so we need the way most, we need the cruises, even with some of the setbacks we've seen, on the cruise side and, even with the, the horrific behavior in San Francisco this last weekend with the Waymo vehicle, we still need those, but I think we also need the OEMs to step up and really push to get the, in Mobileye terms, the eyes off, hands off capability, at least in some ODDs. And they need to architect the autonomy platform, which is the vehicle and smartly sets the requirements.

And I think they need to not do that in a vacuum. They need to do that in conjunction with some of their key tier ones and other players in the field. I would love to help, and Spartan would love to help in that as well. And then with that, we could then deliver that capability in a cost-effective manner.

One of the phrases I've said for a while is like autonomy moves at the speed of the architect. architecting the autonomy system will set the pace of it. And I think right now, and we were talking about this in multiple folks at CES recently, the OEMs are playing a game of chicken, like as long as no one else has all four.

We don't need to have all four, right? But as soon as somebody has it then that's going to be a game changer. And I think that's what I would like to see is like having us having that aspect of a push, keep the way most of the cruises, they're going to do their parts of the market. But we, but I would love to have OEMs really step up and help us push this for our society.

Grayson Brulte:

Hopefully they will, and I'm going to turn that on here. You mentioned architect, architecture is great. So we're going to call you Matt, the architect. Now, Matt, the architect, if you were to build an SCE level four autonomous vehicle today, what sensors would you put on it? And why? 

Matt Markel:

First of all, I wouldn't make a robo taxi, as way more crews they're off near they're off doing that, but for me and the reason I'm starting with the answers that will affect the sensing modalities that we put on there, but On the Robotaxis, I have a really tough time closing the gap on the economics.

And the and making it ubiquitous enough to make a difference in safety. So you can close the gap in the economics. If you just stay in San Francisco, assuming that this last weekend was a, was an anomaly because there's like a, gazillion rides taken. It's something short of 600, 000 miles are driven every workday there via ride hailing apps.

Some ridiculous amount. And San Francisco is, it's small. It's like roughly the key area. There's five by six miles. So the. And it's got really good weather, but the economics get sketchy when you start going to bigger areas. If you want to have a high-quality ride. So high quality ride means you have to have lots of vehicles.

The wait time is low. Lots of vehicles is expensive. And then the number of deadhead miles to move people around in the big areas is really takes a bite out of that out of the economics of that. So instead, I'd put it in cards that you own, and I'd select the ODDs that allow you to incrementally roll out the capability.

So with that, now, to answer the question, the sensors you'll need microphones for certain things where, you know, siren detection and so forth like that. And the microphone people, they actually have a, it's not easy to do the microphones because it's all the other noise and stuff out there and where do you put them and wind noise on the vehicle and stuff like that.

That's not super easy, but that's probably not the main, just the question, the main, we're getting like, lasers in or out, radars in or out. So the, so for me, it's the cameras and the radar. I think, and that's you've heard me talk about the, and I've used this example a lot because I really love them, the Reese's Peanut Butter Cups, the, how the chocolate and the peanut butter go together just so well.

Radar and camera are, do that same thing. Where one suffers, the other comes in and saves the day. Radar, we talked about suffering in spatial resolution. The even with Spartans clarify software that makes it better. It's nowhere near what a camera is. When you get that, 3000 or so, times increasing carrier frequency, the resolution just can't help but go up.

But cameras suffer in certain types of weather where radars don't, and cameras obviously don't have range, and radars do, and cameras don't have velocity, and radars do, and cameras aren't an act, actually an active sensing mechanism. They are, they basically rely on recognition of passive signals in the image plane, right?

Which could be a reflection, it could be a, it could be something actually there. Whereas Radar and lasers to for that sense are more of an active something emanates from the sensor goes out and reflects back and it's only there if the if there's something truly in the scene and to make it, full L4 capability.

You. You need to be able to not have the driver be able to take over and not require them to take over. But you can. But you can constrain the ODD and I think that's the key right there, is you constrain the ODDs and you roll that out. Buy a software over time as you get. More areas and you get more capabilities.

And I think that to me is the key. And that's how we can get that out there. I think I said a while back that if I had true L4 capability, meaning I can, I don't have to, no eyes on, eyes off and in a car today, even if it was just a freeway. I'd probably leave this interview right now and go out and buy it.

And then apologize to you later for leaving the interview early because it would be that much of a game changer. And anyway, so that's what I put on there. I'm big on the camera and radar because it's also the way that we're going to get the price to be something that everybody can afford.

Grayson Brulte:

You're right. Is camera radar, is that the ultimate scaling for an L4 system purely from a, let's say from a performance standpoint and from an economic standpoint? 

Matt Markel:

I think from a performance standpoint, more sensors is going to be better, but I think that doesn't necessarily mean that More sensors are necessary, right?

And I think that better is often mistaken for, we mistake faster for better. So I think it's one thing to say, I'm going to, and this is what Waymo did. I'm gonna throw a lot of sensors on there. And even if I don't need all of them, I will, I can be more sure that I can get to where I need to with these sensors.

And then I'll figure out perhaps how to take them off later, and of course It's difficult because you architect the system on a certain plane with all of them there, and you really take a different type of mindset to say, if we get rid of the crutch of laser, how much does it really affect things?

It's a way to get there quicker, but it's also, it can become a crutch. But yes, the answer is I think radar cameras. Are the right sensing modalities. And I think that's the way that it's going to get out there from a performance standpoint is the way it's going to get it out there on the cost standpoint.

Grayson Brulte:

Could 4D radar technically or hypothetically replace LIDAR, do you think at some point in the future? 

Matt Markel:

In some use cases, yes, I think in other use cases, it's going to it's, one of the things we've talked about architecture a lot as you architect the solution, for the particular problem you're trying to solve, then the laser might be the right thing.

I think there are some indoor applications where in Freight moving around and things like that, where laser might be, since it doesn't have to worry about the weather and, the laser may be a useful alternative there. But I think that the, if you think about the L4 autonomy, we've become accustomed to the crutch of the laser point cloud.

And we haven't really pulled all the information out of the visible spectrum that we could have, and we haven't pulled the information out of the radar spectrum that we could have, because we're, because the laser just looks so good, at least in the situations where, when it does it's not in fog, it's not in, it's not in dust and things like that.

Grayson Brulte:

Well, a lot's going to change. We, the thing that's going to make all this work is cost. If it's not, if it's not cost effective it's not going to scale. At CES 2024, you and Wipro, you demonstrate how a combination of improved camera, radar, and software can achieve SAE level four autonomy. Could you talk about that partnership, please? 

Matt Markel:

Yeah. So Wipro is, they're fantastic and they're investors in us. They've been, we've demonstrated some really cool things together. They are, the relationship is we're Wipro's partner at the application level for their cloud car ecosystem. So they have this vision for software defined vehicles.

So not just the radar, but the whole vehicle software defined that focuses on architecting cost effective 8s and autonomy capability and radar is absolutely critical to their vision for that. They really, because of the cost, they don't see LIDAR having a place in mass market. And they see radar is central, radar is critical.

Radars, need, needed for autonomy, but they don't see a place for lasers. And that's what makes the relationship really compelling. compelling to us and we're working with them on. So how do we work together? As we take this cloud car software defined ecosystem, as they take this to the different customers, we're able to joint sell into them and provide their opportunities for us to provide clarify in our perception stack as well as because of the integration of that and their middleware and overall vision, it's a way for the whole thing to look more attractive.

If you went to buy an iPhone and there were no apps for it, it'd be like less of a compelling purchase. But now with us, we have the, the, you have the iPhone from them and the things like Clarify our resolution enhancement software as the, as one of the apps that they can get as well.

Grayson Brulte:

It's smart because it's scalable at the end of the day. 

Matt Markel:

Certain parts of the world where we've got really good connections, there's certain parts of the world where we don't and they do. So it's a great way to cover the entirety of the planet just with their deep connections. And they're huge. There are billions and billions of revenue every year. So it's a great partnership to have.

Grayson Brulte:

You mentioned earlier about commercial vehicles and the use cases for radar. I want to go one step down that road and go into ag cause ag is really hot. Deere today announced that they're going to build new electric tractors in North Carolina, fully made in America, which is really neat.

And there's a growing movement and Deere's been very public on this around increasing autonomy across. The agriculture CNH is doing it as well. Could radar potentially help increase the performance of the autonomous tractors? You're on a controlled ODD and you're going at what? Five miles an hour, six miles. So you're not going very fast. 

Matt Markel:

I am a absolutely huge fan of John Deere. And that is a not only just a great, American story, but a tech, a company that is so deeply entrenched in the technology development, as well as, all the other things that they have to do also, they have to make, these heavy machines that are reliable and durable and so forth.

But to be so absolutely focused on technology development, it is amazing. I get to spend some time with them this last year and I'm a huge fan. I I think that they're just, they're approaching things the right way. So as they look to radar, I don't see Radar doing the heavy lifting of See & Spray because of the, because of just some of the difficulties with that.

I think that's much more geared towards than something in the optical spectrum, but because Radar can see through dust really well, that it is a lot of, which is not the limiting factor for a lot of commercial vehicle applications. It's a real asset. And I think there's a couple of ways that can, off the top of my head, that can play out one Is the radar can tell you like where you are and what's around you.

And it's like the, we call it slam, the simultaneous localization and mapping, radar can be, it can be super useful with that as well as, for like overall optimization of the route, as well as route correction along the way, and radar, what we see in data from, from our testing is we routinely see things like curbs and other things. And I think that as you see berms and plants, you'll find that as, as well, even at the, even in the RF spectrum. So the as opposed to just, needing a laser or something like that.

And then there's the safety aspect, both of the machine and its operator and of course all those around it. And that's the concept of like free space estimation is similar to what we were talking about before on the L4. It's is something there and what I expect to see.

See something and, for agriculture, that's really important for mining applications. That's, that's really important as well. Here, am I about to go off go off the edge, the is there something there? Should I expect to see something, and of course, from like the safety standpoint, seeing animals, and people in the scene, I think is super important, the, and because, Of the, again, the environment and the dust and so forth, radar typically works better there than lasers and some other sensors, and when you fuse it properly, can be a great pairing with cameras.

Grayson Brulte:

On a farm, one of the most dangerous jobs is the tillage with the hay in doing that, 'cause a lot of individuals fortunately, lose their arms, can radar see through cornfield that there's an object there or something in the hay that might cause damage to the tractor? Can you get just a, can you see a mass to that farmer would know that there's a potential issue there? 

Matt Markel:

I think often, yes. It obviously the ultimate engineering answer of it depends, comes into play but there's the ability to do. To see through and see, have the radar waves. Radar, radio frequency waves are really interesting in the way that they can, that they move around things and creep around things.

There's a whole lot, whole bunch of stuff that the electromagnetic engineers really get into on how these waves move and so forth. So typically, you might say ah, you don't have clear line of sight, you can't see anything. That's not going to work, but actually there's between reflective paths and then and then just working, diffraction paths and so forth it's got a lot of capabilities there.

So obviously it depends on the situation, but there's a lot of capabilities that I think we could bring to bear there. 

Grayson Brulte:

Radar's cool. So you got off road, you have on road, you have military DOD. What are the opportunities for radar in general? Is it just pretty much vast? You mentioned maritime earlier, or is there any limitations of where you could take the technology?

Matt Markel:

It doesn't work so well underwater. Now we have sonar for that, but it's but other than that, there's, I'm continually amazed at just how the different things you can do with radar and the different new things that we're coming into the, obviously, there's a zillion defense cases, but there's things like perimeter sensing now that we're seeing more and more use of radar.

Okay. There's a ton of stuff in counter UAV, counter drones that are important, not just from a military standpoint, but from a commercial and civilian physical security standpoint. And then there's whole other fields of things like the heart rate sensing, breathing, monitoring, sleep monitoring.

We've, I think we've talked to the past about in cabin monitoring, but also just the, just for using it in, at home and the, to supplement in the states here, we're very much looking in detail at all these attributes that go into our health. How are we sleeping? What's our heart rate variability?

All of these sorts of things. And radar is one of the sensors that can do that from a predominantly a non-intrusive standpoint. You don't have to be wearing something for it to work a lot of times. So I think that's a, that's another attribute as well. That's, it's a little different than, Autonomy for vehicles and maritime applications and missiles and so forth.

Grayson Brulte:

But overall, do you see the radar market expanding over the coming decade? 

Matt Markel:

Absolutely. Absolutely. There's As you can as you think about the not only those applications, but other ways to bring Mostly cost-effective solutions to problems. I think you know radars just continue to grow I think it's one of the key areas it's going to grow in is You Fortunately, in the commercial vehicles, bringing that technology, all the technology, all the good that's gone into passenger cars as one of the things that Spartans doing right now is we're, our products for that market are really trying to bring the goodness that we have in, in our passenger cars today, even though it's not full autonomy yet.

Bringing that level of goodness into commercial vehicles across the board, whether that's on road, off road, forklifts, everything to bring that level of safety and capability to them. 

Grayson Brulte:

I want to highlight commercial or give a partnership with Philips Industrial, which is really for the trailers. You're going to save a lot of potential lives. You're going to, you're going to save a lot of time and you're going to overall increase efficiency. Why focus on the trailers from a commercial aspect?

Matt Markel:

When you look at the application these guys are in, and how they're having to back up to different things, and where the, and where people can get injured, and people can be, can get hurt, it's not an obvious use case, but for those that are actually in the space, it's, it becomes obvious.

And that's why things like Philips is developing something called Rear View now, which is the camera only version of that. And what we're working with them on is the next gen of that, which will be Camera plus radar, which I think is really, I think that there actually several folks are maybe even wanting that more than just the camera only version, because you do get so much more from the redundancy of the sensing modalities and the separability of the spectrum. So that's what we're providing there is like that camera plus radar, but when you think about that, so that's, so right. So trailers is one, one use case, but there's a ton of, we've already talked about like how we radars and camera together, right?

So the, you can think of like lots of other opportunities for this product, which is going to be, it's going to be cost effective. It's going to be produced by Phillips at their, at their world class production facilities. And it's gonna be super solid. It's gonna have, our technology and their technology all in this.

There's a lot of other applications as well. So again, it's not just the trailers. It's a yes. And trailers and a bunch of other things as well. 

Grayson Brulte:

It's your Reese's Peanut Butter Cups moment. Yeah, you got it. 

Matt Markel:

It's exactly it. Yeah. It's the Reese's peanut butter cup right there on the, coming first to the back of a trailer.

Grayson Brulte:

So Matt, in your opinion, what is the future Spartan Radar?? 

Matt Markel:

So this is actually a really super exciting time for us at Spartan. We're definitely selling at scale this year. We're. We're crossing the chasm, as they say on revenues, primarily focused on Hoplo, which is our radar product, but also it's big on the R& D side.

So Hoplo is our, as we talked earlier about what we're doing for commercial vehicles, that's under the brand name Hoplo that we're selling that. That radar system. And so big on the sales side for that, but on the R&D side, also there's, we just mentioned what we're doing with Phillips, we're doing, this is the year of the engineering and integration, hoping to be in production by the by the end of this year for that product, we're also working on additional software features for Hoplo, which is we have the base set right now, but again, because it's a software defined radar, we can continue to add things over time and increase the capabilities of that, including a new way to sell this, where fleets can purchase the hardware and the software separately. And then they just download and install the latest version of the software. Also special configurations of that software, perhaps geared towards them.

So if you think about One of the things about the industry right now is, if you want a radar that goes in the back, you buy the back radar. If you want a radar that goes in the side, you use the side, you buy the side radar. Since Hoplo is software defined, we can have the same piece of hardware support all of those, and then we just instantiate it with whatever they want.

Back side with static object removal maximum detection range they want, number of zones, number of segments. Thanks. And then other special things like, do you want the mining overlay on top of this? Do you want the, whatever, as we continue to develop these. And be able to sell this in a way that they, that is optimized for that particular for that particular OEM or distributor.

So they, they get exactly the software they want at the time that they want to. And then on the automotive side, we're continuing to refine, clarify the eyeglasses for radar to. Continue to improve speed and performance as well as looking beyond that. And what does the clarify enhanced perception stack look like?

And how can we really make that a product that we bring to market as well. So big year for sales but also a really big year for the engineering team. And to be honest, that's perfect for me. That's exactly where I want to be bringing products to customers to solve their problems and then doing the cool engineering to keep that pipeline going.

Grayson Brulte:

Engineering makes the world go round. It's the most incredible things are built by engineers. The whole world's built by engineers. Just everybody has their little niche. And they build really cool things and you're building really cool things with the team at Spartan Radar in the radar niche. As we look to wrap up this insightful conversation, Matt, what would you like our listeners to take away with them today?

Matt Markel:

Grayson, thank you so much for the time here. Whenever I get a chance to chat with you, it's super fun and really informative, and this was just a great conversation. And realistically, with all the topics out there in the automotive space, I want to thank you for devoting the time to deep dive on sensors and the sensor business.

Obviously, I'm very passionate about it. Sensors are the key enablers for safety and basically everything we want to do in the mobility space, whether it's passenger cars or light trucks or commercial vehicles over the road or off road. And just thanks for raising the awareness on this.
I think that's super important. I still believe that radar is the most underused sensor in the automotive in automotive applications today. And, by continuing to focus on things we'll take a bite out of that. Just, I ask is just for the listeners is just keep informed.

You can follow Spartan Radar on social. We're on LinkedIn. You can connect with me on social, email me at matt.markel@spartanradar.Com. And while you're at spartanradar. com, you can check out our website, look at the different products there. I also write a letter from the CEO. I was going to do this every month, but it's it turns out it's about every quarter where I highlight an area that needs to be raised and is worthy of some different discussion some further discussion of things.

So not a sales pitch or anything like that but the from my view of the world, what's the, where are things going and so forth. So feel free to check that out. And yeah. And just stay informed. 

Grayson Brulte:

Thank you. The letters I read them, they're well done. I always learn something.

And this, I'm going to highlight this again. The most underused sensor radar, a lot's going to change. I'm really curious where the market goes with radar. Today is tomorrow. Tomorrow's today. The future is Spartan radar, Matt. Thank you so much for coming on SAE tomorrow. 

Matt Markel:

Thank you, Grayson. Have a great day.

Grayson Brulte:

Thank you for listening to SAE Tomorrow Today. If you've enjoyed this episode and would like to hear more, please kindly rate, review, and let us know what topics you'd like for us to explore next. 

SAE international makes no representations as to the accuracy of the information presented in this podcast. The information and opinions are for general information only. SAE International does not endorse, approve, recommend, or certify any information, product, process, service, or organization presented or mentioned in this podcast.


Listen to the full back catalog on your favorite podcast platform.