Delete the human
No crew, no problem
In today’s Diary:
A mattress on the factory floor
1,000 miles the wrong way
Dig once, bury forever
Driving in Vegas, steering from Texas
Tunnels on Mars
The first time I visited David Zagaynov, he was sleeping in a small room above his factory.
Downstairs sat his first aircraft, Seagull, a 13-foot unmanned “bird” built to take off and land on water:
When I returned to Poseidon‘s warehouse in the old shipyard neighborhood of San Francisco a year later, the floor was crowded with boat molds. Engines were firing up on test stands, roaring through the building.
They’d gone from one small plane to three full-size aircraft in the time it takes Boeing to schedule a design review.
Poseidon builds cargo planes with nobody on board, not even a pilot. And it just raised $60 million to transform one of the most important, and most backward, markets in America.
It’s one of two innovators I’ve visited recently who are taking the human out of the machine.
Planes are just machines to keep humans alive at 35,000 feet.
A pressurized tube so we can breathe. Heating because it’s -70° F outside. Oxygen masks… emergency exits… little lights along the floor.
Airframes are engineered at ruinous expense so if the worst happens, we fragile lumps of flesh inside might survive. Certifying a new airliner runs into the billions.
Making flying safe is one of the great unsung achievements of the last 50 years. But a parcel doesn’t need any of it.
Roughly $8.3 trillion worth of goods is shipped by air every year. And almost all of it moves inside machines built for people.
Many of the world’s big cargo jets are passenger planes that got old and were sold to freight companies. Pull out the seats and the overhead bins and start loading pallets.
People are heavy for the space we take up. Parcels are light but bulky. So a tube shaped around people fills up with boxes long before it gets near its weight limit.
That’s why many cargo flights are full of “air” and partly why shipping parcels costs an arm and a leg.
Poseidon said: “Screw that.”
Licensed pilots will still fly Poseidon’s planes... from the ground.
What comes out is the human on board and everything needed to keep them alive: the cockpit, windows, pressurization and seats. And out goes the rule that the entire structure must survive a crash.
Take out all the padding and the airframe gets lighter. A lighter plane needs less thrust. Less thrust means a smaller, cheaper engine. A smaller engine burns far less fuel. Fewer systems mean fewer parts, and fewer parts mean less maintenance.
You also delete the clock.
A pilot can fly about eight hours a day. That’s the law, because tired brains make mistakes that kill people.
A plane with nobody on board lands, loads and flies again. Which means you can multiply the number of hours every plane can work. As David put it: “The plane doesn’t have a family to come home to.”
I think Poseidon wins because it’s allowed to break things.
A founder I met this year told me about a Lockheed vice president who said to him: “Do you know how far ahead we’d be if we could crash things?”
SpaceX won partly by blowing up rockets and learning from the wreckage, which you can only do with nobody aboard.
Rip the human out and you earn the right to fail… which is the right to get better, fast.
A Cessna Caravan, the workhorse of regional cargo, costs $1,100 to $1,200 per flight hour. Poseidon is targeting $400 to $500, about 60% less.
A cheaper plane transforms where freight can go.
If you’ve ever had a box arrive overnight, thank Fred Smith.
In 1973 he launched Federal Express and built it into one of the largest logistics companies on earth.
To fly freight between cities you used to need a plane going directly between them. Twenty-five cities meant 300 different routes. Fred only had 14 jets.
So he sent everything to one city in the middle of the country, Memphis, sorted it all in the middle of the night, and flew it back out again. This meant a package from Boston to New York flew about 1,000 miles in the wrong direction, got sorted at 3 am, and flew back north before dawn.
Fifty years later air cargo still runs on his idea.
David brought up Fred Smith before I did. “Air cargo today is hub and spoke… and it’s not the most efficient or the best.”
Fred’s answer was right for Fred’s planes.
But cut the cost of an hour in the air by 60%, let it fly around the clock, and suddenly it makes sense to just send the box directly to where it’s going.
This is point-to-point shipping, which David calls the “holy grail” of logistics.
FedEx and UPS will be his customers. The losers will be the small regional airlines who fly those short routes today.
Meet Egret…
Poseidon’s land plane is made for those short runs. Its wings stretch 50 feet from tip to tip, and there’s nowhere inside for a person to sit:
Egret carries around 3,250 pounds of cargo. It takes off and lands in a short distance, so it can use rough gravel strips instead of a paved runway. It has cargo doors at the front and back, a ramp that folds down, and a flat floor a forklift can easily load onto. Somebody finally designed the plane around the box!
If you know anything about how container ships transformed global trade, you know how groundbreaking that is.
Egret’s first flight is targeted for the end of the year.
Then there’s “Heron,” the seaplane. Same size, about two tons of cargo, and 1,600 nautical miles of range:
They share the same wings, tail, electronics and engines. Only the body is different: one is shaped to land on water, the other on dirt. And the wings come off so the whole plane packs into a shipping container. It can go anywhere a truck, train or ship can take it.
What you won’t find on either is anything fancy. Just two ordinary combustion engines, the kind mechanics already know how to fix.
I can tell from talking to David he’s obsessed with dodging the #1 killer of hard tech startups: they build one amazing machine but can’t build 100 more.
That’s why Poseidon’s factory looks like a boatyard.
Traditional planes are riveted together from thousands of metal pieces. Poseidon lays sheets of carbon fiber into enormous molds and sets them into big single pieces.
A Boeing 747 contains 6 million parts from more than 550 suppliers. Poseidon’s airframes need a small fraction of that. Every part you don’t need is a supplier who won’t let you down.
The first production line is being set up to build two to three planes a month. Then David says: “We’re going to 10X that.”
Those planes are designed to go places no cargo plane can, including islands, coastlines and remote towns that today are either a slow boat or an expensive flight away.
That’s why the military is interested. Poseidon already tested its smaller plane, Seagull, with the US Navy and Coast Guard.
Poseidon just moved into the 110,000-square-foot hangar where America made aircraft in WWII.
Three years ago David was still a software engineer at Amazon when he noticed every part of its logistics machine had been hyper-optimized except the planes. He asked his roommates: “Why can’t we build planes?”
Now Poseidon is ushering in the biggest change in moving freight since Fred Smith sent everything through Memphis.
Take the pilot out, and the airport shrinks to a runway and a warehouse. David has floated building his own network of cargo airports, with no terminals and no security lines. And Heron won’t even need the runway.
Packages could hop between smaller warehouses like trucking works now. His dream is a network so fast that a shipment is “not even worth tracking. It’s like you press the button and then it’s right there.”
Getting to Hawaii used to mean more than four days on a ship or a long, expensive flight in a propeller plane. It was a place for the wealthy. Then in 1959, Pan Am put Boeing 707 jets on the route and cut the flight to under five hours. Tourism 10X’d in a decade.
What gets created when Poseidon slashes the cost of moving cargo around?
East of Austin, TX, a small town called Bastrop has become America’s frontier tech hub.
On one side you have SpaceX’s giant Starlink factory. On the other is The Boring Company, which I think is Elon Musk’s next trillion-dollar company:
Elon got the idea when he was fed up with LA traffic.
We stack thousands of people into 50-story office towers. Then at 5 pm we send them all home on roads one layer thick. No wonder there’s traffic! We need to make our cities 3D by building down.
There’s one catch: it only works if tunneling gets at least 10X cheaper.
Picture a steel worm about as long as a football field. That’s a typical boring machine.
Up front a giant wheel spins and grinds through dirt and rock. Conveyor belts carry the dirt out, and a mechanical arm lines the walls with curved concrete panels one ring at a time.
A big one costs upward of $80 million. And when the tunnel is finished, they toss it like a used coffee cup.
Boring machines are custom-built for one job. Hauling them out would cost more than they’re worth, so they’re left in the ground. Dead boring machines are under cities all over the world.
And you can’t get better at something you only do once.
Make something over and over, and it gets cheaper. That’s why TVs, solar panels and batteries keep falling in price. But almost every tunnel is a one-off, so tunnels have stayed expensive.
Musk’s companies have a habit of finding a cost everyone treats like a law of nature and knocking a zero off it.
At SpaceX it was the cost of lifting a kilogram into orbit. At The Boring Company it’s the cost per mile of tunnel:
The Boring Company takes Elon’s playbook underground.
Instead of a custom machine for every tunnel, The Boring Company builds one reusable machine called Prufrock.
Prufrock arrives on a truck, tilts nose-down and dives straight into the ground. When the tunnel is done, it digs back up to the surface and right onto a trailer the company calls “The Monster,” which drives it to the next job.
Seeing Prufrock up close is magic. I’ve watched one surfacing, and it’s like watching a giant mechanical mole:
The steel wheel at the front is studded with cutting discs that spin, crack the rock, and grind it to pieces while water keeps the dust down. Then the broken rock drops onto a conveyor belt that carries it the length of the machine and out of the tunnel.
A normal boring machine digs forward about five feet, stops, and builds a ring of concrete to hold up the walls. Each ring takes 20 to 45 minutes.
Prufrock digs and lines the tunnel without stopping. Boring makes the curved concrete panels ahead of time. Inside the machine, a mechanical arm fits them into rings behind the cutting wheel while it keeps digging.
The team is pushing to fit each five-foot ring in under four minutes!
And this giant steel worm is one of the cheapest parts.
Out of every $100 spent digging tunnels, $35 goes on design, reviews, permits, management and financing. Another $25 pays the workers.
More than half the cost is people and paperwork about people. So what happens when you take humans out of the tunnel?
That’s what The Boring Company is finding out with Zero People In Tunnel (ZPIT).
In Bastrop there’s a room with desks and screens. While I stood there boring machines chewed through desert dirt under Las Vegas… steered from Texas! Even the trucks carrying concrete panels to the machine are driven remotely by a guy with a PlayStation controller.
“Delete” humans from the machine, and you slash tunnel costs.
New York’s Second Avenue Subway cost about $2.5 billion a mile. The Boring Company’s first system, two tunnels and three stations under the Las Vegas Convention Center, cost less than $30 million a mile and only took six months!
In Las Vegas right now you can walk into a station, hop into a Tesla, and ride under the city straight to where you’re going.
This is the Vegas Loop, The Boring Company’s first deployment. Today it has 14 stations stretching from the convention center to Strip hotels and out to the airport. The Loop has carried more than 4 million passengers.
Las Vegas built a $2 billion stadium for the Raiders. But after a game even the high rollers in the luxury boxes sit in traffic for two and a half hours because 65,000 fans are all trying to squeeze onto the same roads. Once the Loop reaches the stadium you’ll be out in under 10 minutes.
My advice to any mayor reading this: letting The Boring Company dig is one of the cheapest, fastest things you will ever do for your city. Say yes and say it quickly.
Last week The Boring Company raised $3 billion in a round led by the United Arab Emirates, my new home country.
Work has already started on the Dubai Loop, and the UAE deal could eventually mean more than 150 kilometers of tunnels across the country. And Elon isn’t stopping at Dubai.
Prufrock is exactly as wide as a SpaceX Falcon 9 rocket.
That means parts can be swapped with SpaceX. But a digging machine the same size as a rocket can also ride in one. See where I’m going with this?
Mars has no air and nothing to stop radiation from space. One of the best shields is a thick layer of dirt over your head.
On Earth digging with nobody in the tunnel saves money. On Mars it might be the only way to dig.
Unmanned boring machines on Mars. Sounds crazy. So did rockets that land themselves.
For 50 years planes stopped getting faster and tunnels stayed absurdly expensive. After touring dozens of factories across America this year, I can tell you that’s changing.
The innovation famine is well and truly over. Our job as rational optimists is to make sure people see it.
Share this Diary so more people follow the builders making the future great
—Stephen McBride
PS: My first book, Disruption Investing, is officially live on Amazon. Thanks to all who pre-ordered… you made it the #1 overall bestseller in stock market investing, which I am truly grateful for. Now that it’s live, you can get it on Amazon here or by clicking the image below.
