Loading summary
Transcript1 lines
- [00:03]
A
My first thought when I heard about Mana was that drone delivery didn't work. Just look at the tech giants Amazon and Google in 2013. Then CEO Jeff Bezos announced on the CBS TV show 60 Minutes that Amazon was testing deliveries with drones. A year later, Google Alphabet subsidiary Wing started up their drone work too. Yet after a decade, delivery, especially in the United States, is still done by a guy in a truck, and nothing about that seemed to be changing anytime soon. So I guess I felt that if Google and Amazon hadn't pulled it off yet, with all the resources available to them, then there must be something inherently wrong with the technology or business model. But then I listened to Ben Thompson's interview with Mana CEO and founder Bobby Healy. Bobby sounded like he knew the business rather well, and so I decided to make some time to come to Dublin to do a visit. After five long years in Europe, the drone delivery industry is starting to take off and shifting regulations open a window for the same in the US in the coming months. In today's video, I flew to Ireland to visit a drone delivery startup and they answered every question I had. I want to note here, by the way, that Manna didn't sponsor this video. I paid my own way to Dublin. It's a pleasant place, though quite gusty. A challenging nursery for a drone delivery company. Dublin's climate is a fine example of the ongoing challenges of weather prediction. The day's weather will waver from sunny to windy to rainy to cloudy to sunny to windy to night. And yes, it drives me crazy. So on Friday morning, I arrive at a Dublin suburb near a McDonald's. The winds are strong and it is rather chilly, though city regulars might say the weather was quite pleasant by Dublin standards. As I get off the car and walk towards the McDonald's to get something warm to drink, I see out of the left corner of my eye a drone taking off from behind a nondescript fence. I cannot help but stop and watch as the drone rises vertically into the air, slowly turns to orient itself, and then speeds away. Bobby shows up a bit later. He is youthful, a charismatic fellow with a fine beard and a nice Irish accent. If you squint and he is wearing his glasses, you might imagine that you're talking to kind of like a Steven Spielberg. That's just my own vibe. Mana isn't Bobby's first startup, but it is his first one doing hardware. Having done this before, he is a confident leader, but also unusually open about how Mana runs. Perhaps it is his tactic to get us to believe it actually works, but Steen is actually believing Mana's chosen bread and butter is fast food. Their most frequently ordered item is a cup or two of hot coffee, a small, relatively light order where much of the enjoyment is tied to how soon you can get it. These are very much impulse buys. The focus on fast food in the suburbs is strategic and tied to the current bad economics of food delivery in the suburbs. Someone in the suburbs wouldn't often order delivery from McDonald's or Starbucks because the delivery costs too much, about $9 to $11 per order. There are a few reasons for this. The obvious one is that the suburbs are more spread out. Drivers must drive longer in the suburb, maybe 10 to 15 miles as compared to 2 to 5 miles in the city. In the most ideal scenario, a Delivery takes about 20 to 30 minutes on average, depending on traffic. But oftentimes the scenario is not so ideal. For example, waits at the restaurant. Restaurateurs often prioritize, drive thru and dine in orders or just get crushed by mealtime rush hours, leaving the delivery guy waiting up to 10 minutes to pick up the food. Even with time padding added on by Uber or DoorDash, they aren't getting paid for that wait. So with that, a 20 to 30 minute round trip can easily balloon to 40 minutes, translating to a per driver throughput of about 1.3 to 1.5 orders per hour. Base pay after Uber or DoorDash takes their share is about 3 to $4, which at 1.5 orders per hour translates to 5 to $6 earned, which is nowhere near enough to pay car expenses and gas. So American drivers particularly rely on a good tip, raising earnings and what the customer pays to about 9 to $11. Delivery drivers in the city suffer similar issues like long restaurant waits and bad traffic. But in urban areas, the economics are somewhat better because drivers can batch orders and drive a smaller, cheaper vehicle like a scooter or E bike. This lowers their cost and raises throughput, meaning cheaper delivery prices for the urban customer. But in the suburbs, batching is infrequent and most drivers have no choice but to drive a car. The end result is that customers in the suburbs are charged more for food delivery, so most prefer to cook at home or drive themselves to the restaurant. Manna has multiple drone bases in this particular suburb, known as Blanchardstown. They first entered the area in February 2024. This particular base is sited in a fenced off corner of a McDonald's parking lot, has five drone pads for takeoff and landing, and is serviced by an operations trailer. The trailer more like A converted shipping container is a unit of work, as Bobby likes to say. It houses operating equipment, control systems and the battery charging racks. The trailer will make it easy to open or move a new base. Just haul it over and plant it down. This flexibility is important because the base's real estate footprint is a critical part of the drone delivery business. Finding new locations can be challenging. When the customer sends in the order, the restaurant starts making the food. Meanwhile, Manna might start shuffling drones around, like sending one over to cover a particular location in case of a shortage of delivery drones. Once the food is ready, a human runner is sent to fetch it from the restaurant. A human loader then loads the food into the drone. Meanwhile, Mana prepares a flight plan to the customer. The customer must manually choose and confirm via their app a suitable location for drop off. Behind the scenes. The software quickly calculates a flight plan and submits the digital paperwork for the flight to what is called a utm, which stands for an Unmanned Aircraft System Traffic Management. The UTM approves the flight and then communicates with other UTMs in the area to deconflict the path. Finally, the control panel shows a green button on an iPad like tablet next to the door. The operator begins the flight by pressing and holding said button and and then off they go. The key 10x advantage of drone delivery is simple. A drone can fly in a largely straight line to the customer at about 55-60 mph, bypassing traffic, traffic lights, construction, what have you. Flight times are suddenly slashed 10x and become far more consistent. But drones change other parts of the old equation. The fewer batched orders you see in the suburbs becomes an advantage because one batched orders mean longer waits at the restaurant, which means it takes even longer for the customer to get their food, especially if they're last on the batch chain, and 2 it gives you simpler point to point flight paths. Moreover, and this part differs depending on the particular suburb and its planning. Most retail outlets and restaurants are co located in places like strip malls, so a drone base can quickly access most all of the shops in a relatively central area. The key question is how much human labor is being expended per delivery. For Mana, those are the human runner bringing the food from the restaurant, the human loader putting food into the drone, and the human pilot in the control center overseeing the drone. There remains room for improvement the by automation, particularly the runners and loaders. The former is especially time inefficient and probably needs to be automated in the future, perhaps with something like a robot car. Considering that the Business model pairs quite well with dark kitchens, locales that only sell meals for delivery. The food can be made at a central location and then quickly routed over to a drone. Bobby emphasizes the positives of not having food delivery drivers on suburban roads, not only with regard to their car's potential carbon emissions when applicable, but also safety. Food delivery drivers are constantly on the road, going to unfamiliar places and tend to be rushing. It's not their fault they're trying to make a living, but such speed, unfamiliarity and carelessness and leads to incidents. Statistics are rare, but surveys indicate that at least a quarter of riders drivers suffer an incident or accident of some kind whilst working. All in all, I buy that after inputting the test order, we get into a car and drive about eight to 10 minutes to the delivery site. Mana notes that they often deliver to driveways or backyards for single family housing, but they can do taller apartment buildings too, either by dropping on rooftops if the customer has access, or a courtyard. They're already doing that for their trials in Helsinki anyway. The drone is so fast that we must push the order down in the order queue, otherwise it will have already been delivered by the time we arrive. Once we are in position, we signal the team to go and send the drone. Bobby opens up a cool internal app showing the drone heading down its authorized flight path. The drone arrives in two to three minutes after it launches, emerging from behind the tree line at a rather high speed. You wouldn't really see it coming unless you know where it would be coming from. The sound is like a quiet burr, like as if someone is flying a strange propeller plane above. Once it reaches the delivery zone, it hovers over the location and slowly descends to about 46ft, or 14 meters altitude. Then suddenly, a little door opens at the bottom of the drone and you see the bag of coffee drop towards the ground on a string at about 4 meters per second. Five seconds later, the payload softly hits the ground. The drone cuts the string on its own and then heads back on its merry little way. The whole process is amazing to watch. It really works, and impressive considering that it was relatively windy, at least by Taipei standards, and the drone has to deal with wind shear from the nearby buildings. Bobby said that the drones can fly in winds as high as 40 miles per hour, and over the past 365 days in Dublin, they've flown 97.5% of the available time. Mana tells customers that they should not try to come out and get the food while the drone is dropping it. Pilots won't do the delivery if they can see people below them, they've seen customers try to chase the food, try to cut the string themselves, and so on. Please don't try this, they say. Just wait and watch from a safe distance. Anyway, the whole thing happens fast. Less than a minute later, the the drone is gone and the food is left sitting in front of your house or your garden or backyard. The coffee was still piping hot. Drone Delivery's customer experience is killer. It can give food to you fast enough that it's still good to eat. Hot food delivered via scooter gets shoved into a silver box and there inside it waits for 20, 30 minutes while the rider putters about sweating and getting not so tasty. French fries are the classic example. French fries are wonderful the moment they come out of the fryer, but soon after that they start to get soggy and sad to eat. This only compounds the guilt you feel whilst eating something so unhealthy. And of course they can deliver more than just food. You can also get things delivered like phone chargers and USB cables, things you often need right now but are so small that it makes driving down to the convenience store feel like a waste for chore. One thing that initially surprised me, but also sort of made sense was books. Wanna start that new novel from renowned author Dan Brown right now? Get it dropped on your backyard. The Mana headquarters is a nondescript two story building in a Dublin industrial area. On the first floor is the workshop where the drones are assembled. There I saw a number of drones at various stages of work in progress. If you've seen before the drones that Google and Amazon are working on, then Mana's drones might seem exceedingly simple. 8. Propeller's motors on arms attach to a carbon and aluminum frame with sensors and antennas. A plastic cover with UV coating goes on top of the whole thing. If desired, someone can brand the drone too. In the back there is a parachute. This is the safety device designed to go off if something ever goes wrong. Out of the company's 200,000 plus flights, this happened just once. But having it there adds two magnitudes of safety for people on the ground. That's it. Simple and frankly a bit ugly. This simplicity was a deliberate choice. However, simple means reliable, less things to go wrong, and Again, it's the SpaceX School of Design. The best part is no part. The drone's control systems are also very simple and engineered for triple redundancy. The first thing that jumps out at you are these three orange boxes. Each box contains a CPU that receives data from the sensors and outputs control signals for the motors per its programming. A separate module weighs the three CPUs votes and and chooses the majority vote. A backup battery sits on the board able to provide power for just enough time to do emergency procedures. This is all borrowed from the civil aviation world as like the rest of the drone's design. On the drone's bottom is a single camera so that pilots back at the HQ can use to monitor and review the drone's food drop off and landing processes. The parts are all off the shelf and with the exception of the motors which come from mainland China, are sourced from within Europe. Inside the control room you can see pilots guiding the drones. Per regulations these pilots have to be trained, but since the flight path is predetermined and the drones largely run themselves, the pilots main job is to play the role of overseer. This way a single pilot can oversee about 60 to 80 deliveries per hour. When things go wrong, it is the pilot's job to abort the flight and pop the onboard safety parachute. This floats the drone slowly down to the ground. The most critical time is at the drop off. The pilot looks around through the downward facing camera to ensure that nobody is standing under the drone when it drops the food. Some drone delivery companies have engineered their drones to descend down to the ground to deliver the package. This is not ideal because below a certain height the parachute is no longer an effective safety tactic. So you have to engineer more into the design leading to knock on effects that complicates the design and balloon's cost, complexity and time. Better to avoid this, keep the drone simple and just use the string to rappel the food down even if the thread costs $0.15 per drop. One thing I wanted to learn about was how Mana generates demand. These drones are capital investments like $150 million EUV machines from ASML. They are rapidly depreciating. We need to be utilizing them all the time. A drone sitting on the ground waiting for orders is not making any money. For demand, Mana partners with food delivery aggregators like Wolt, Deliveroo, Doordash and Just Eat with more on the way. Mana does run their own delivery app, but it doesn't seem like that is their core strategy. Bobby says that doing direct to consumer is a beast that requires a lot of advertising, operations and above all cash. Better to just work with the aggregators. Now I have a little experience with robotic SaaS, businesses companies with robots that sell their services to enterprise customers and one of the first demands that prospective customers make is exclusivity. They want the robot all to themselves as a competitive advantage. If an aggregator really wanted it, they can get it, but they would have to pay extra for it because exclusivity reduces volume and that hurts the overall economics of the business. What customers in Ireland and elsewhere are saying so far is that price really matters. Mana's costs are about half that of human delivery right now, but they want to eventually scale that down to less than a dollar per delivery, cost per flight and volume throughput are the keys to getting there. Moreover, there's the issue of real estate. There are only so many ideal launch pad locations, especially near a restaurant or strip mall, since it is often those guys real estate that the drones are taking off from. Though not at this particular base. Mana prefers exclusivity with the restaurants themselves. One of the first things that I wanted to know about was safety. Are these going to be safe for people? Europe and Ireland in particular has been working on its drone regulations for years. There is a lot of lingo to process, but the key idea is is about risk mitigation. If you fly a drone over populations of people, then you must explain to the regulator how you are keeping those people from harm. You must engineer multiple levels of safety and redundancy into the drones and the procedures. The EU has a risk assessment methodology that roughly speaking leads to levels of drone operations safety called specific Assurance integrity levels or sail. There are six levels of increasing risk from sail 1 to sail 6 based on population density, drone size, the amount of risk mitigations and so on. Cell one is flying a drone where nobody is around. Lesser need for mitigations. Cell 6, on the other hand is where a drone carries human passengers in the city, like what is being trialed in the People's Republic of China. Since a catastrophic failure with that means very bad things will happen, you must engineer the drone as like an Airbus. So like how certification standards are set up such that the probability of experiencing catastrophic failure conditions is less than one in every billion flight hours. Mana's drones are at Sail 3, which is a moderate risk. This is because the drones are unmanned and fly over relatively sparse populations, so they can be engineered to a medium robustness. About one catastrophic terminal velocity type outcome for every 10,000 flight hours. If they do crash, then their small size and light weight means the collateral damage won't be as bad as something heavier. And as I mentioned earlier, the onboard parachute adds two turns of risk mitigation, bringing it up to 1 per 1 million flight hours or 8 million flights. Add the fact that the flight routes are programmed to avoid highly populated areas like stadiums, and that means that the risk of a MANA drone falling out of the sky and hitting someone is very low. This is not a mana only risk mitigation framework. This is what the EU has been using since 2019, drawn from civil aviation and over the years that drone and drone delivery framework has proven its efficacy and the industry has avoided bad accidents. Though people regularly acknowledge that one will eventually happen, it is simply inevitable. The hope is that the industry will have matured enough by then to survive it. So all in all, Europe has done a good job in handling the safety aspect of this civil aviation regulation, but provincial and local issues remain, the most prominent of which are community inhabitants, concerns about privacy and noise. With regards to privacy, the worry is that the drones are videoing people in their homes and backyards whilst en route. The drones do indeed have a video camera on their underside, but it is only used for food drop off and delivery. I know you have to take them at their word when they say that, but it seems easy enough to catch them on it. Next is noise. If you do a generic Google search for Mana, Wing, Zipline or any of the other drone delivery services, some of the first results to come up discuss noise, perhaps using ominous language in doing so. Look it up on Hacker News and you see several highly upvoted articles asking how do I prevent drones from flying over my home? Mana acknowledges that noise concerns exist. I can hear the drones burr as they take off and land at the launch pad. Here it is. When the drone passes overhead. However, this sound lasts for just about 10 seconds and is measured at about 53 DBA DBA, meaning that the sound measurement is weighted for human hearing. To get a better sense of how it actually feels, here's what it kind of sounds like. When the drone descends for the drop off, the sound increases to about 62 DBA for about 30 seconds. Here it is again. Let me bring up a few comparisons. 60 DBA is about the level of normal speech and is lower than a vacuum cleaner or roadside traffic, so statistically speaking it is not unprecedented. But it is a new and different kind of sound, so people will likely be more sensitive to it. To further reduce the noise, the MANA team is rolling out a new version of the drone with larger propellers. They are about 40% quieter. Moreover, their drone routing algorithm has heat polygons on the map that warm up the more times that a drone flies overhead, so the drone will try to avoid areas it has recently passed over. So all in all, I reckon the noise will not bother most people, but the ones who will get bothered are going to make this issue their life's crusade. Community protests and lawsuits will ensue. A third concern, a more American one, concerns someone like a neighbor or a bystander shooting down the drone with a shotgun. When en route, the drones are too high up in the sky, but become somewhat more vulnerable when they descend to drop off the food. Mana was threatened as such during their Dallas trial, and it has actually happened at least once to another company's drone, apparently by a customer's elderly neighbor. But shooting down a drone is a federal offense, same as trying to shoot down a regular plane. The FAA even has a small FAQ on their website saying this shooting down a drone creates debris that can cause harm to the people below on the ground. So don't do it. The question that myself and many other people, investors included, have wanted to if the economics are so compelling, then why hasn't it taken off yet? And what in the United States means that things are about to change? Bobby has a few thoughts on this. First, Mana has spent the last five or six years getting the operations right and proving as such to the regulators. They are essentially starting a small airline and it was not until the last 12 months that the company has really gotten its training wheels taken off, regulation wise. Second, the US regulation situation with regards to drone delivery has has been unfavorable. For all the complexity of the EU regulation scheme, the EU's big upside is that their system is a waiver based permissions regime. This means that the EU puts out a set of regulations that apply to everyone. So once compliant, a company like Mana can easily enter the rest of the European market, conceptually at least on the ground. Reality is of course complicated. It was different in the United States. The US Federal government had a permissions based regime which meant that each drone player had to get approval from the FAA on a case by case basis. Without that, drones had to be within visual line of sight of the drone operators. So if the drone had to travel three miles to deliver something, then operators had to put three people on the path to keep the thing within sight. Then in early August 2025, the US government put up new rules called Part 108. All in all, it follows Europe's direction in moving to the waiver system. The FAA will create a framework of broadly accepted safety standards and a drone delivery company that complies with those standards is granted permission to operate across the country. This rule is not official yet. Right now the FAA is still collecting feedback, but if it progresses as it is expected to then it is a major step forward with regards to the US drone delivery industry. The doors are swinging open. Dublin has been a great playground for drone players from Amazon to Mana to Zipline. The problem however is that Ireland hosts just 5.3 million people and that is just not a big enough market. So the future of companies like Mana is overseas. As of this writing, Mana has done about 250,000 flights in total and they are planning to scale up to 2 to 3 million flights in 2026 and then 20 to 30 million flights the year after that. And what I have seen thus far says that the model can scale. Mana's expansion plans include cities in Germany, the United Kingdom and more. One interesting coming locale is the United Arab Emirates. The UAE is a great example where the drones superiority in customer experience really matters. I heard it gets hot there, so drones really can help. There are also other intriguing locations like Mexico City. The urban core is not really drone delivery friendly, but many of the city's 20 million people live in the suburbs. With traffic jams quite horrendous, drone delivery can do good there. But of course the United states and its 70 to 90 million suburban households is the world's most prized market hands down. The big question in my mind is whether or not Mana is going to be allowed to be there. One major outstanding issue is the nationality question. Right now only American drone delivery companies are allowed to operate for commercial gain in the United States. This is not a recent Trump thing, but dates back to old civil airline industry rules called Kabataj. Only American based airlines can pick up and drop off customers and cargo at two points inside the us so from SF to la, Cabotage has been around since the earliest days of the commercial airline industry and has remained despite deregulation elsewhere, perhaps due to fears of foreign subsidized airlines like Emirates. So long as Cabotage applies, Mana can't pick up cargo for pay. But they seem fairly confident that the US Department of Transportation will eventually change, especially since the EU's drone delivery market is open to American players like Google and Amazon. Moreover, Mana is willing and able to build drones in the US too. What governor would want to pass up that photo op Once that is settled, another competition. Major competitors like Google Wing are already gearing up in response to the changing regulation environment. Wing's expanding partnership with Walmart seems to hold much promise. If there's one non tech giant company that can bring volume, it's Walmart. And VC is a funny thing. They are momentum investors. Once the first American drone delivery player hits 10 million revenue in the United States, then we can anticipate VC money funding a flood of new players. It will be a rush to grab territory and sign up vendors. MANA will hopefully be participating in that. I don't see why they can't compete. Ryanair is a clear example of an airline that kicks butt, but even if they do not enter, it seems that their footing is solid in a very fruitful EU market. I want to thank the MANA team for accommodating me, and based on what I saw and heard from them and others in the industry, it looks like drone delivery is finally starting its rise into the sky. Alright everyone, that's it for tonight. Thanks for watching. Subscribe to the channel, sign up for the Patreon and I'll see you guys next time.