theCUBE + NYSE Wired: Physical AI & Robotics Leaders | Ashish Parikh, Doodle Labs

Clips
More from theCUBE + NYSE Wired: Physical AI & Robotics Leaders

Speakers

Ashish Parikh

Co-CEO

Doodle Labs

play_circle_outline Revolutionizing Robotics and Drones: Enhancing Systems with Jetson and Qualcomm Integration

play_circle_outline Rapid growth at Doodle Labs, hiring intelligent people for robust data links.

Info
Transcript

Ashish Parikh, Doodle Labs

Ashish Parikh

Co-CEO Doodle Labs

In this session at theCUBE's East Coast studio during the NYSE Wired Robotics and AI Media Week, Doodle Labs' Ashish Parikh, co-CEO, explores the forefront of advanced wireless communication systems crucial for autonomous drones and robotics.

Parikh, an expert in radio systems, explains how Doodle Labs designs and implements communication networks for robotics, emphasizing their importance in creating reliable and robust links for autonomous operations. The discussion provides insights into the complexities involved in connecting autonomous mobile ro... Read more

explore Keep Exploring

What are important factors for the best drone companies to consider in order to have successful operations and connectivity for their systems? add

What kind of growth has Doodle Labs experienced in the past five or six years in both robotics and AI? add

bolt Powered by CUBE AI

Ashish Parikh, Doodle Labs

search

>> Welcome back, everyone. I'm John Furrier, host of theCUBE. We are here live at our NYSE studios on the East Coast. This is where all the action happens on the East Coast. Our subnets, our point of presence, our access point, bringing tech, money and culture together. Of course, theCUBE studios in Palo Alto is where Silicon Valley ecosystem is. That's our access point there, connecting Silicon Valley and Wall Street. Money and tech, all in one network with the NYSE Wired community, it's an open network of trusted individuals developing very, very fast, sharing their knowledge here on theCUBE and at the NYSE. Ashish Parikh is here, co-CEO of Doodle Labs. Welcome to theCUBE. Thanks for coming on.

Ashish Parikh

>> Thanks so much for having me.

>> Can't wait to get into what you're working on because wireless technology that you're doing is fundamental for a lot of the robotics that are out there. We hear self-driving drones, autonomous drones, unattended retail, supplies going from point A to point B, all kinds of humanoids, all ought to be connected somewhere. So, you guys are working on this. So, first of all, let's get into what you guys do. Set the context, what does Doodle Labs do?

Ashish Parikh

>> Doodle Labs builds the radio systems that robots and drones use to communicate with each other. So, for example, some of the examples you were even talking about. Let's say you have a warehouse with a whole bunch of autonomous robots moving around, in and out between shelves. Our radios are used to create the wireless network that they can all use to communicate with each other and with the human operators that are overseeing the entire operation.

>> So, give an example of what it's used for, because I mean, getting on the same page. This communications between devices, there's a lot of technology in robots. We've been unpacking that all week. The computer vision's critical, the software-hardware integration, but they also have to collect data and then also talk to other data sources, but also, talk to each other and understand what's going on in the system.

Ashish Parikh

>> That's right.

>> Does it have to be a use case that's well-defined? Does the blanket coverage have to be already there? Take us through some of the specifics of how that system works.

Ashish Parikh

>> That's a great question. There's two ways to think about connectivity. There's this general sense of connectivity, how things can move around in these very general purpose types of cases. But as you were mentioning, robotics are becoming more and more specialized. You could take a general-purpose device and then you want to customize it to do a specific function. So, going back to my example about picking things on a shelf. There's very specialized types of requirements about how they move around, how they avoid banging into each other, what they're doing to collect. And therefore, when you have these specialized robots, you also have very specialized needs for what the communication is going to look like. So, let's say that one of these robots moving around a warehouse goes offline, the entire warehouse often will have to shut down to make sure that this robot can be brought back online, so that things don't collide with each other. A bunch of safety issues. So, the communication system is of utmost importance to stay super, super resilient.

>> On the unmanned mobile robots, AMR is an acronym, I've learned that it stands for autonomous mobile robots. You got vehicles, you got factories. Is there an environment you guys don't do right now? What do you guys have right now and what's coming or what's not available? Separate where that connectivity layer sits and what's the enablement?

Ashish Parikh

>> Where we focus on our use cases, let's call it, less than 200 kilometers in distance. And what we specialize in are what we call private networks. So, let's say that you have a group of drones flying around. Typically, they're not going to be on some public network. You're going to have to create a private mesh network between them, and that's really where we specialize. So, if you have your own system where you need some really robust connectivity within the system, that's what we specialize in.

>> And give a use case. Give an example.

Ashish Parikh

>> Let's say you have a swarm of drones flying out in a defense use case. Our radios will create a mesh network between these swarms of drones. I mean we've had up to 100 nodes on a mesh network before and they also are able to stream very long distances. So, our drones go over 100 to 200 kilometers in distance. So, you can have the safety of staying at base while your swarm of drones is flying out, streaming video, sending a whole bunch of sensor data back.

>> Talk about the... We all had our experience with our phones connected to the Wi-Fi router stadium, "Oh, I got full bars, but no connectivity." Or one bar. So, I can imagine with the mesh network, you got the drones, they're connecting to each other, so they have to talk to each other, "Hey, are you there?"

Ashish Parikh

>> Yeah. Exactly.

>> "Hello?" And a little handshake happens. And then, now they're operational. You got to have the throughput and the signal. Those are probably some design... Can you share some of the secret sauce behind that? Is that accurate assessment?

Ashish Parikh

>> Yeah, that is actually a lot of the things that we try and optimize. So, it all comes back to the quality of the signal and the quality of the link. The higher the quality of the length and the stronger it is, you're able to send more throughput, more data, you're able to exchange more information with your peers. Whereas sometimes if you're in a, let's say, a contested environment, there's a whole bunch of electronic warfare happening. You have a signal that intentionally needs to stay quiet, then we toggle things to keep the radio as quiet as possible only sending the data that's needed .

>> Yeah, so incognito mode?

Ashish Parikh

>> Exactly.

>> So, signal jamming?

Ashish Parikh

>> Yeah.

>> In a defense scenario.

Ashish Parikh

>> In defense context is called LPI/LPD. And so, you only want to send data out that needs to be sent out, otherwise stay absolutely quiet.

>> On the power side, robots form factors can be large, depending on the use case, or smaller. We've got some entrepreneurs building smaller drones, more elegant. These guys are industrial designers. And so, form factor also gives constraints. How much battery power, power for the radios? You got to propagate the signal. How does that factor into it?

Ashish Parikh

>> It's one of the main design focus of our team is creating things as small as possible to fit in as many of these different use cases as possible. Earlier on your show you had Vantage Robotics on, which we know them quite well. They've released a super small drone, for example. The trade-off there though is that there's a limited amount of range, limited amount of battery, and they've designed their own radio system for that. However, the bulk of what drones out there are doing, they're typically going the 10 to 100-plus kilometer range, and that's where we focus. Our radios, they're typically about the size of a matchbox and the higher power that you add, they do of course grow in size, but they tend to stay pretty small.

>> So, do you guys sell the radio only? Are you selling drones? Take us through what you guys are building.

Ashish Parikh

>> We do both hardware and software, specifically around the wireless network, so the radio itself. So, we have a hardware team that designs radios, and then we have a software team that builds a whole bunch of networking features that layer on top and how they operate.

>> Okay. I want to get that out of the way. So, now people can understand where you fit. Okay, now I'm a developer of a drone. I got the super use case, it's called theCUBE Drone. It's going to go out and find experts with the computer vision, look at someone, look at their LinkedIn, do some doxxing on them, funky use cases and go, "Okay, great. I spotted someone." I got to build that. I got to write the software. What do I do? Am I integrating with you? Is it open source? How do I, as a developer... I have to get the radio. I got to figure out where in the form factor would you be sitting and then write code. Take me through it.

Ashish Parikh

>> You're absolutely right. I mean the hardest part for drone manufacturers is essentially the system engineering that you were just describing. The best drone companies out there are the ones that have the best system engineers and they're able to put all of the different pieces together and then have it operate exactly how they'd like it to. What we do, because we provide all of the connectivity in that piece, so we're one piece of the puzzle that they're working on. We spend a lot of time, one, optimizing our piece of the puzzle, but then we are now spending a lot of time integrating with the Jetsons of the world, with the RB5 and the Qualcomm platforms, and so that our radios are essentially plug and play for the system engineer. You drop it in and it works.

>> I was talking to my friend Jim Long, his son, Jackson, just graduated from University of Illinois. He's working for . He's a system engineer. He said that, "We are living in..." I'll quote him, I'll give him credit for it. "We're living in a world of Lego engineers." He's a mechanical engineer by... Now, he's coding, now he's soldering, he's making circuit boards. So, we're in a world now where the engineering is very componentized, it's very like Lego blocks. You want to build a Millennial Falcon, go ahead. But the parts are there. So, this seems to be kind of the driving force.

Ashish Parikh

>> Yeah. And we even take that even in our world of designing radios, we also think in Lego blocks. So, what we have is like a framework, in which we have a bunch of radios off the shelf that we've taken our own RF building blocks and made things, but when a customer comes to us and says that, "Hey, we have this pretty ambitious project, we need something that can do X, Y and Z. it's a little bit different than what you guys have off the shelf." We're able to rapidly iterate on our products and build customized solutions for them.

>> So, you fit nicely in the landscape, because robotics use you as a piece part, critical piece, radio needs to talk, computer vision, connectivity, kind of the mission-critical subsystems. What other areas of the drones do you guys see in your customer base that become the key subsystems? Computer vision, obviously. Compute, some storage, but compute mainly and radio and networking. Did I miss anything?

Ashish Parikh

>> No, those are frankly the core pieces that go into it. The compute is where some of the most interesting things are happening. Alongside, of course, unbiased on the connectivity piece as well.

>> Without connectivity, nothing works.

Ashish Parikh

>> Exactly. I mean, and also we've all been reading in the news with the Ukraine and Russia conflict about how electronic warfare is the new normal for military use cases. Now, when customers come to us, they don't come to us and say, "Hey, we need a radio that is resilient to jamming attacks sometimes." It's a core part of the requirements that they give to us. Back to this iterative building block approach. We've had thousands of our radios fielded in Ukraine. The feedback we've gotten from the field we've used and iterated very rapidly. I mean last year we put out four separate releases of what we call our Sense protocol, that is an anti-jamming protocol. And so, this is essentially the new normal. And that's why I was saying that there's a lot of innovation happening just in the communication side.

>> Back in the day they called it software-defined radios. Software's a big part of it. What are some of the things you're seeing on the innovation side with software? And where's the commercial advantages now? Defense, clearly a lay up, but let's shift to commercial. Where's the software advantages? How is commercial developing? What are some of the areas that you're seeing develop?

Ashish Parikh

>> So, you have your hardware and then what your software can do is optimize how that hardware works given a set of limitations. So, there are actually a number of limitations out there on the commercial side, especially for example, in Europe they have extremely strict power limits. So, you can't even send very strong signals to do long distance that drones require. So, what software can do is, back to when I was talking about the quality and the type of data you send through, you can optimize your link with what you have available to make it robust for specific use cases. So, there's a lot of work that we're doing about optimizing our links to make them really resilient at long distances or when there's a lot of noise and interference. I mentioned that Sense protocol about jamming avoidance, we have a version of that that is used just for regular interference avoidance. So, in here there's a million Wi-Fi nodes in this room, for example, a lot of interference happening. Our software will find the best channel, the cleanest channel for a radio to operate at. So, it can still-

>> And that's software. And consumers, we have ear pods, they change tunes sometimes based upon what's the hearing. So, this is really where the advancements are coming. Okay, so the next question from that is, okay, where's the advancements coming? What are you guys seeing that's happening on the entrepreneurial side or ecosystem side around you guys? Because you got the enablement, you got the core subsystems and building blocks with software. Where's the innovation coming from?

Ashish Parikh

>> So, the innovation, I believe is still around this topic of being able to pass data and have an operational system in really noisy or jammed environments. Number two, it's about deeper integration. You referenced compute. There's more and more compute, there's more and more autonomy on drones and robots now. And so, there can be more and more of a link between the two of the computer informing when do I need to send data? Why do I need to send data? In what manner?

>> All right. So, tell a story. Stories drive movements. Wireless is a story that's part of robotics, of course, AI. A story that's cool around what you guys have done. The cool robotics use case of commercial or defense, and then how AI is infusing in here to make it go faster. Go faster in terms of the trend. Not so... Oh, make the drones go faster too. That'd be great too, but-

Ashish Parikh

>> Probably a very tangible example for listeners could be drone delivery, which is very soon coming, guys. So, let's just paint the entire picture. Let's say you have a Walmart with a whole bunch of drones flying out from a Walmart. So, that you're going to have a whole bunch of drones in the sky and they're going to be dropping packages out onto your lawn. So, here, the wireless connectivity is pretty challenging. You have a whole bunch of drones that need to be able to communicate with each other, let them know where they're moving to. Just like when you order an Uber Eats package, there's a photo taken when the package's delivered, that needs to be sent back to your app. There's a whole bunch of data moving around. So, where AI fits into this is about essentially informing the data links, or even in the broader system context, optimizing how the whole network moves around. This is coming really soon.

>> Yeah, and that's congestion and contention for airspace on drone delivery.

Ashish Parikh

>> That's right. And then, you're also navigating a whole bunch of FCC and FAA requirements about where you can and what type of signals you're allowed to send and when and at what power levels.

>> Talk about the company, what's the buzz? What's the exciting things around Doodle? People are watching. I know a lot of people are moving into the space. New talent, customers. What should they know about what you're working on? What's the cool thing? Put the plug in. What's going on?

Ashish Parikh

>> Doodle Labs, we're fortunate, we're riding some really amazing waves both in robotics and in AI in general. We've been growing at a very rapid pace, 50% year-over-year for the last five or six years.

>> Where are you guys located?

Ashish Parikh

>> We're in Los Angeles and we also have an office in Singapore. Our team is growing very rapidly. We're bringing on some really, really intelligent people as we're building more and more robust data links. If I were a young person looking to join the robotics space, I would be looking at getting my hands dirty as fast as possible with some part of that system, and I think we offer that.

>> Yeah, just jump in.

Ashish Parikh

>> Yeah, absolutely.

>> I mean, there's a lot of growth. There's a lot of new challenges. What I find with tech people, certainly in our network with theCUBE, the alphas, I call them the alphas, they want to work on hard problems. I mean, plenty of hard problems to solve right now. What are some hard problems that you guys see that if you throw it out there as kind of bait, if you will, recruiting bait or whatever you want to call it. I mean, people want to know what are you working on? Is it the backend? Sending the picture back... I mean, there's a lot of backend process problems. Certain databases work here. You got a data problem, you got data lakes, connectivity. What are some of the hard technical comp-sci, engineering, software problems that you guys are trying to tackle this year?

Ashish Parikh

>> Defense tech is really hot right now, as we all know. And one of the main reasons for this is there's this paradigm shift of there's this huge wave of autonomous and unmanned vehicles that are going to be forming the future of our military and the militaries around the world for a lot of really good reasons. Really big problems here though. Now, as we're trying to transform our military from something that was... You have humans driving flying planes and following pre-planned missions to now thousands of drones that need to be controlled, that need to be able to communicate with each other, that need to be able to send information back to the humans that are navigating the entire operation. Massive, massive challenges. How do you have a swarm of 100 drones talking to a group of individual soldiers that are down on the ground over here in this country and something that's happening in a continent over there? There has to be really, really strong communication infrastructure behind all of that.

>> Yeah, I was hanging out on Chatham House rules with the General Mattis crowd, and learned that 43% of military deaths are from friendly fire. And drones, technology solves that problem at the tactical edge. I mean, think about that. 43%, almost half the deaths.

Ashish Parikh

>> That's a statistic I didn't know.

>> It's around there. I might be off by a few points, but it's greater than 30%. It might be 43% if I remember correctly.

Ashish Parikh

>> It's a large number though.

>> It's a huge number.

Ashish Parikh

>> And it's an amazing thing to be able to remove humans from the situation. I mean, look at what's happening in Ukraine and Russia. Drones are driving the entire offensive and defensive posture of the whole thing. And there've already been thousands of deaths there, but think of how many more people would've died if it was just soldiers running up to each other trying to shoot each other. Hopefully, this has a huge, huge change .

>> Certainly, there's going to be a balance of power. Ashish, thanks for coming on, appreciate it. And Doodle Labs doing great work. Again, love the networking conversations. Networking connects. Without connectivity and power, nothing exists. Congratulations on the success.

Ashish Parikh

>> Thanks so much for having me.

>> We'll have to visit you in LA sometime. We'll certainly continue the conversation. I could talk about this for hours. Friday, it's end of the week, of Robotics Week and AI leaders. Again, part of theCUBE and the NYSE Wired community, an open network of people sharing their knowledge and just connecting. I'm John Furrier, host of theCUBE. Thanks for watching.

Ashish Parikh

>> Thanks so much, John.