Bobby Healy, Manna | theCUBE + NYSE Wired: Physical AI & Robotics

Hello out there, studio connecting Silicon Valley and Wall Street. >> I'm John Fortt, and this is here with Dave my co-host. >> Welcome back to the Cube studio here at the New York Stock Exchange. I'm Gemma Allen, co-host of NYSE Wired Physically AI series. And joining me now is a man from a location very close to my heart. One of Ireland's, if not Ireland's, most successful tech entrepreneurs, Bobby Healy. >> Thanks for having me, Gemma. >> Welcome, Bobby. >> Great to be here. >> So, let's get straight into it. You have recently relocated Manna air delivery from Ireland to Tulsa. >> Tulsa, Oklahoma. >> Oklahoma. >> Yeah. >> How did a boy from Kilmacud decide, you know what, I am going to hire a thousand people in Oklahoma? Give me the backstory. >> Very simple. We already have 220 people in Dublin, where we do our R&D, our manufacturing, and we run our drone delivery operation. Uh USA is the biggest market that would be interesting to a company like ours. And because thanks to strong regulation here and changes moving forward with the regulation, aviation regulation, the market's now fully open. So, USA is open for business, and so this was the right time for us to really start focusing and investing in the United States. And I go where the action is. I'm the CEO of the business, so I have to be in the USA. And we chose Tulsa, Oklahoma as our first manufacturing base and our operational base. And so I'm delighted to be there. And I'm really happy to be there, actually. >> Okay. So, let's get into the actual company. We actually talk a lot about drones on NYSE Wired, but not typically from the perspective of consumer delivery, right? We talk about drones in defense tech and physically AI, and you know, AI on the edge. Your business, though, is very much in the consumer space, right? >> Yeah, very much. >> So, talk me through it. Like, you are promising to deliver food, anything I guess really FMCG, in a much shorter space of time for a much lesser cost. >> 60 miles an hour in a straight line, your burrito will arrive in 3 minutes. If you live in Tulsa now, you can order your burrito, it'll arrive in your back garden in about 3 minutes, perfectly hot, lands gently, and when you it's mad as it sounds, a flying robot with AI is is 100 times more efficient than a person in a 3,000-lb car driving your burger. So, if anyone's a DoorDash or New Breed customer, you know, regular customer, you know the pain of not knowing when your product's going to arrive, your french fries slightly cold, maybe some of them missing. Um drones don't eat french fries, so it's a much better experience. >> Yes. >> And we've been flying for 6 years in Europe, and we've an NPS between high 80s, early 90s. People just love the product, but physically physics is on our side, and the cost of operating a drone to do that delivery is a fraction of the cost of a human in a car. So, if there's no way you're it's the only way you're going to get your products delivered in the next 10 years in the United States. >> So, physics on your side, talk about the financial side of that. So, let's say you get a burrito in New York, it's probably $25 for a burrito, God knows. In Tulsa, I hope it's a little bit cheaper. Right, so let's say $15. >> a steak in Tulsa. >> What are the economics? Like, break it down, and who takes the cost? >> That's a good point. So, primary the primary reason we exist is cost. So, the cost of labor of a road-based delivery is about $12. That's what it's costing to make that happen. So, that cost is passed on to the consumer in surcharges and fees. It's a it's a necessary evil. And no one's doing really well in that three-sided marketplace. The restaurant are not doing great. The aggregators, you know, struggle to make margin with with higher labor costs and labor rules going in the wrong direction. So, the answer is automation, the answer is robots. Our cost today, even subscale, and we're only like 400,000 deliveries now, so it's subscale relative to the aggregators, our cost is already a fraction of what it costs to make a road-based delivery happen. And as we scale, our costs are going to come down. So, you you'll see a world easily in the next 5 to 10 years where the marginal cost for road-based delivery, last mile delivery, will be in the region of 50 cents to a dollar. And so, imagine how that impacts on demand and consumer behavior, purchasing for food, getting parcels, going to the hardware store. All those things will reorganize themselves around a fully digital automated last mile system. >> Let's talk about the Ireland situation. I think we have to cross that situation. Right, because I am Irish, too, and I follow the Irish news. So, I know there were some regulatory challenges that made you decide to explore the US. >> Yep. >> Huge market here, right? But also, I'm sure a more competitive market. Like I know DoorDash have said that they're going to potentially create their own drone infrastructure. You have Amazon. I mean, you know, the 600-lb gorilla, right? How do you think about it from the perspective of, you know, like a proof of concept which obviously went very well at home, to now taking on the giant? >> At home, we had the largest drone delivery operation in the world in Dublin. >> Wow. >> And unfortunately, EU is great at regulating things, but not so great at making things happen. Um so, we struggled to get planning permission. We struggled with local city regulations. It's just taking too long. But it's kind of doesn't really matter. I mean, our business was always created to build this technology. It's a technology business. We have our aircraft now. It's well tested thanks to our and thanks to strong aviation regulation in Ireland. Um but when it comes to local city regulations, it's a different story. Uh whereas the USA, it's open for business. I mean, Phoenix zoned the entire city for drone delivery. Where we are in Oklahoma, it takes us 2 weeks to get permission to fly. Texas, we're flying. We've been flying for years in Texas. And you see already by 10 million deliveries in Texas between the top four players. And we're in the top four. So, there was never going to be a question of us trying to grow in Ireland or Europe. USA was always our market. It was just a matter of timing when the regulations, the aviation regulations, opened up. Uh which they have now. So, look, you know, if I were building this business again, I'd probably still start in Ireland because we have great technology base, really strong government support, and we're backed by the sovereign wealth fund of Ireland and Enterprise Ireland. So, we have really strong government support, but just when it comes down to the nitty-gritties, you it's difficult to do business in Europe. It's fragmented, it's complicated legal situation with planning, with regulations. They're different in every country, and the USA's one single gigantic market. And to the to the competition, they're all trying. Like, Google have a drone delivery program, Amazon have a drone delivery program, Zipline have a drone delivery program, but they're we're all in neck and neck, and it doesn't like there's 6 billion deliveries that need to happen last mile to replace the existing industry. So, I wish them all the best. You know, I'd I'd I'd happily share it with all three of them, and we'll still all be billionaires. >> I love it. I've always loved your confidence, Bobby. But, back to the point you made, the quantity there you're you're quantitatively the largest drone delivery or consumer drone delivery service in the world, right? What's that measured on? Is that measured on a percentage of consumer spend? Like, how do you actually create those numbers? >> It's measured on the hard part. So, it's measured on So, viable unit economics. So, we have profitable unit economics where we once we make money on every flight, first thing. That's it. That's table stakes once you get past safety. And then the hard part is throughput per operational hour. So, in that one location, Blanchardstown, I'm sure you know Blanchardstown very well. We were doing three or 400 deliveries a day or eight deliveries per aircraft hour. So, the way to model the success of a business in this space is the way you would model a low-cost airline. What is the turnaround time? What is the productivity of that aircraft? What's the turnaround time? How much revenue per passenger are you making? How much revenue a day are you making per aircraft? That's how we measured ourselves. Measured ourselves not in cool or sexy or any of that stuff. It's the the bare metal, how much money are we making per aircraft? And nobody could touch us with that. And nobody can touch us with it cuz of our architecture. >> Well, let's get into the tech then. Let's go under the hood of these drones, right? >> So, you were a Nintendo developer by >> I'm a yeah, programmer. I wrote some of the world's worst video games, yeah. Michael Jackson's Moonwalker, play it, it's great. >> But, you're a tech guy at heart. Okay, a lot is happening in this world of physical AI. We talk a lot about AI on the edge here, too, on the show. I mean, it's like a really growing conversation, right? When we look at the actual competition side of it, though, right? From my perspective, oftentimes, a lot of the spec kind of sounds the same, right? Like, you know, folks are building, you know, on Vera Rubin, they're building, you know, cheaper commute for like AI on the edge, physical AI, right? Like, we hear all the same spiel. >> Yeah. >> How do you typically though or how do you tactically and honestly compete on the tech? >> So, every So, the drone is not the is not the reason you're going to be better. In fact, the drone's almost irrelevant. Once you get the drone, it's safe, you can manufacture it at scale, and and and your operation around it works. We're all the same. Like, I I hate to put myself down, but like really, it's not about the drone. It's about the operation that runs the drone. How many deliveries per hour can you get? And like our number, very simple, we get eight deliveries per hour per aircraft, and everybody else gets less than two deliveries per hour per aircraft. And that's not deep tech, it's not AI. It's a very efficient operation with an aircraft that doesn't need to charge when it's sitting on the on the deck. So, if you look at everybody else, they charge the aircraft while it's waiting for next flight, and that takes half an hour. Um so, we've got a process just like a low-cost airline would, where the aircraft's on the ground, it's back in the air in 20 seconds with new cargo, and it's flying in a straight line. Our air our airspace management software routes everything intelligently. Their AI is used, but around more weather prediction and what the wind's going to be over there 2 miles away or 3 miles away. That's where we use a lot of modeling. And but but most of it is heuristic space, so it's just simple, here's the rule, here's the process, get the product into the aircraft as quickly as possible, and get the aircraft up in the air. And our number there is 20 seconds, and I would say nobody can touch us with that. So, it's it's what Michael O'Leary would say is, get all the costs out, you know, pay people well, and have a very solid safety record that lets you scale the business. And that's what's important. So, you know, when when we're doing a million flights a day, and and we'll be at 2 million flights annualized by the end of this year, when you get to that level of scale, the law of large numbers works against you. So, if you're not ready, if your aircraft isn't mature, and hasn't lived in the trenches for years, you're going to have problems. And and so, to to anyone that's going to start building a drone delivery program now, I applaud them, and I think they should do it cuz it's such a super industry, but it's going to take years to get to that scale that you need to get to really be a viable business. And we've been flying for 6 years. We're probably the most match-fit company in the world. >> So, this is essentially like a super efficient data plane that kind of keeps these drones running at maximum efficiency, right? Like, >> Yeah. >> all the time. How easy is that to replicate? Like, how easy is it for AWS or Amazon or Zipline or Google Wing or whatever to wake up tomorrow and go, "I'm going to do what Bobby is Heli's doing?" >> Well, they are. I mean, so Amazon have a super drone delivery program, and they've stated they're going to be in 500 cities over the next 3 years. Uh Wing, Google Wing, Alphabet, uh have a absolutely superb drone delivery program, safe, ready to scale, a great team behind it, and a great balance sheet. And and so, Zipline similarly, they're they're all great companies. I wouldn't say anything bad about any of them, um but it's thin air, right? There's 6 billion road-based deliveries annually, growing at 15%. Nobody is going to be doing that type of volume anytime soon. So, actually, what the customer wants, and to us, the customer is the aggregator, the customer is Uber Eats, the customer is DoorDash, Just Eat in Ireland, and so on. So, for us, it's about empowering those aggregators to get products to their customers in a just a better experience way, and lower their cost. That's what it's about. And for us to take 10% of the market, it's going to take us 5 years to 10 years. There's so much green market ahead of us of to capture that shift from road-based delivery to air-based delivery. I wish them all the best. And in in fact, we talked to each other a lot. >> Let's talk about the partnership model then for a second on the actual business model, right? Because one thing we know about wholesalers like Amazon is they're also vultures, right? Like, you know, for the most part that's how consumers view them, okay? Like they take And the small businesses especially view them that way, right? They take a big percentage. That's no no slight on Amazon, it's just the model, right? Like across the board. A lot of companies never went direct to consumer for web because they didn't really have the infrastructure in order to delivery, like the logistics, right? Like that is a huge challenge. >> Yeah. >> So, who is in scope for this from the perspective of businesses that could potentially use drone delivery to expand? >> Everybody. >> I think a lot of our small cap, mid-cap businesses here in the US that have gotten left behind by so many technology wins, right? >> If you've got a flat roof or a one car parking space available, you can have drone delivery for your business. And that drone delivery will reach someone 5 miles away in 5 to 10 minutes. So, instantly when when we give you drone delivery, suddenly you've got 200, 300,000 customers within a 3 to 5 minute flight. So, that's going to empower small businesses to grow and to go D2C. And if you look at what we're doing in Tulsa, we're working with Pariah Coffee, right? A small little trailer making coffee. And we doubled their business because now they can reach 150,000 customers by air instead of the customers to walk up to their site. And that's a tiny little business, but imagine the hardware store, the bookstore. Like our bookstore beats Amazon for delivery times already and that's subscale because our logistics is distributed. The logistics is all over the city, so you can be a business anywhere in Tulsa city and avail of drone delivery and and beat Amazon because Amazon have the guy in the truck, right? So, Amazon's benefit and and and the way they win is scale, right? Economies of scale through a logistics system, and that's what lets them do such a great job. And drone delivery is the opposite. It's a distributed system that's more democratized that every small business can get access to. So, it's going to be a gigantic job creator for those small businesses in cities. >> How easy is it for them to deploy this on their website or their kind of delivery app or whatever it is, right? Is this very plug-and-play? Like we also know that things like Stripe and all of these other technologies aren't not are not necessarily as easy for small businesses and mid-cap businesses as others, right? >> You have a one-page contract, sign the line, the drones arrive, done. >> What Okay. >> And they're in our shop or they can have their own shop, and we give them an API, two lines of code. They hit our API, so at the end of their shopping basket, our code goes in and says, "Instead of what's your address, it's drop a pin, which is a GPS coordinate where you want the product to land." >> Wow. >> That's it. >> And so, operating this from the first perspective of like, you know, the actual logistical side, it's very straightforward for these businesses. >> For a customer, the end customer, it's super simple. Just get out of the way. Um and and that's it. Choose the pin, approve the pin, i.e., the the destination where you want the product delivered. For the consumer, it's super simple, and most consumers they they never go back to road-based delivery once they use drone delivery. It's not a novel teach, it's the way to get things. For a for a small business, they just got to sign an agreement with us to get the drone delivery. And if they have enough volume there, if they're doing 100 deliveries a day, let's say, we'll we'll house the aircraft permanently on their roof or on the car park. And if they don't, then the aircraft will go there, pick up the product, and go on to the customer from there. >> What's the life expectancy of a drone? >> 75,000 flights. >> What is that like in years? >> Yeah. >> To get a to get a a flying robot to last that long is very difficult. Um and I think that's another area where we would like to say that we excel and probably are the industry leader. Our flights our aircraft are good for 75,000 flights, which means that they're generating significantly more revenue than they've cost us to build. And as our as we scale our cost for building those aircraft will come down a lot as well. So, they need to last forever because if you think about it lasting long for a piece of hardware like that implies that it's reliable. And reliability is kind of important when gravity is the deciding factor on the outcome. >> One last thing on the question, cost right on your end. We hear a lot about total economics. We know that like devices are somewhat different right in the sense of the percentage of commute that they take computer that they take the power efficiency etc. But I'm sure you have a pretty big bill of your own right when it comes to actual software and you know all of the kind of chip infrastructure that goes into this. >> No. >> Do you partner with anyone in particular? Like who how do you build like what what are you building on? >> The hard the biggest cost we have is capex as well on the manufacturing of the aircraft. Once the once the aircraft are built, our cost our marginal cost just goes to maintenance in the aircraft, replacement of batteries. So, the batteries have a certain lifetime, the motors have a certain lifetime. So, we amortize those over their their lifetime. So, for us motors about 1,000 hours of flight, you change the motors. Batteries about 1,000 flights for a battery. So, those are our real marginal cost for delivery. Still like a fraction a fraction of the cost of of labor driving a car. So, we don't again, it's a hardware it's a physically present business. So, you have to think about it like probably a better version of the scooter industry where the unit economics and the capital economics did not work. But for us with a 75,000 flight aircraft, we're just generating a ton of revenue over its lifetime and we'll put those 75,000 flights on the aircraft in probably two to three years depending on the utilization of the aircraft. But there's no token cost there's no you know energy as well. We're using 300 watt hours per flight. So, that's like three 5 cents of energy which we could get from the sun and told us if we wanted to. So, so your actual input cost really are the amortization of the hardware over its lifetime, which is pretty small. >> Sweating the asset. Okay, so you have said that you have will do 2 million this year, 100 million by 2030. I'm not smart enough to know what X that is, maybe 50X? >> Yeah, so 2 million capacity is not 2 million deliveries. For us, that's 120 aircraft deployed in Tulsa. That's 2 million capacity. We'll probably get to probably a million deliveries, I would say, within that by the end of the year. Um I think the way to think about scaling for us is the bottleneck doesn't become manufacturing is easy. These are easy aircraft to build, so if you think about 120 aircraft is 2 million deliveries, it's you know, it's not hard to build 1,000 aircraft. That's 10X growth for 2027. We could do a 50X growth just by manufacturing more aircraft. The bottleneck is actually opening up bases, it's signing real estate deals, it's signing commercial partnerships, it's signing, signing, signing. So it's it's the normal It'd be like running a cable system or or a or a mobile phone roll out, that kind of thing. There's physical presence required, there's real estate required. So we move from being a technology organization to a regulated technology organization by by the aviation regulator, and now on to an enterprise kind of a business where we're making partnerships across the country, and those partnerships will generally be tied to real estate. So we'll sign brand X and we'll we'll fly out of their car park in their 500 locations. So we want to do now is sign the big brands and start taking up that real estate and making those our bases. And then each city, and Tulsa's our first one, each city suddenly gets our mesh network, and that mesh network covers the whole city. And then we do take away food, I mean, that's an obvious one, but there's a lot of other use cases. Parcels is a use case, medicine is a use case, the hardware, bookstore, you name it, you know, it should be flown. And the utilization curve of a drone of of food, for example, it's very peaky in the evenings, weekends, very peaky. So you've got about 60% of your available capacity for the aircraft available for other use cases and the aircraft infrastructure is sitting there. So, we're going to be delivering parcels, we're going to be delivering You name it, we're going to be delivering it. And so, so that's That only becomes viable when you have the entire mesh rolled out in the city and suddenly you can get from any point in the city to any other point in the city in 10-15 minutes. >> Last question, Bobby. I'm getting the signal here from the production team. >> Keep waffling. >> No. I I As always, I'm up on time. What is the priority for you to make all of this happen? Like, are you your own best sales guy? Where will your time be spent here in the US in the next kind of 6 to 10 12 months? >> Yeah. So, we're Right now, we're focused on talking to various states. So, Oklahoma is our first home state, Texas as well. And we're looking for Where's our next six cities that we want to roll out to over the next 12 months? So, that's my job. And then the other part of my job is We're venture-backed company. We've raised $110 million so far. We have plenty of money now. And but at some point in the future, we're going to be wanting to go to that next 100 cities. So, my job is going to be in the United States is building the team here, the leadership team here. You'll have seen we hired Kenny Jacobs recently. So, building a solid leadership team that can expand our operation and just go full throttle. So, we want to be in six cities next year, 50 cities to 100 cities the year after that. And that's going to be my job is driving that growth. >> Well, Bobby Healy, I certainly wish you all the best. Thanks for joining us on NYSE Wired. >> Thanks very much for having me. >> I'm Jen Allen here at the Cube studio at the New York Stock Exchange. This is NYSE Wired, Physical AI. Thanks for watching.