Zipline aims to make a million deliveries a day within two and a half years, after Uber agreed to take all of the capacity Zipline had planned.
Keller Cliffton says Zipline's factory must grow five to ten times bigger than first planned, a shift that makes hardware scale the central bet.
Keller Cliffton argues that hardware companies depend on about 30 disciplines, and a single weak link can sink the entire company.
Zipline bets on a million daily drone deliveries by 2028
Zipline is betting that instant drone delivery becomes a mass-market utility, and Uber's commitment to take its planned capacity pushed the goal to a million deliveries a day within two and a half years. Demand already outruns supply, and customers use the service more once it is easy. The hard part is hardware, since the factory and dozens of engineering disciplines must all scale together.
Takeaways
Uber's demand reset Zipline's target far higher. Zipline planned a million deliveries a day, then Uber's chief executive toured the factory and agreed to take all that capacity. The plan is being replaced, and the factory must grow five to ten times beyond its first design.
Zipline cannot yet meet the demand it has. Utilisation per home ran an order of magnitude above plan, so Zipline deliberately limits homes served and often asks customers to retry later. Keller expects the market to grow another tenfold from about 5.5bn US instant deliveries yearly.
Customers adopt it fast, as with early Uber. Some customers order many times a week, one 80-year-old ordered 350 times in a year, and in some neighbourhoods over half of homes order. Growth ran 10 to 15 percent weekly after full FAA approval.
Hardware means about 30 disciplines, any one fatal. Avionics, firmware, logistics and supply chain must all work, and a weak link can sink the company. Zipline has faced platform two problems, tariffs and maintenance crises that swung fleet health from worst to best within a month.
Zipline sells instant delivery, not drones. The vehicle is only 15 percent of the solution's complexity. Large companies have spent about a billion dollars a year on drone delivery with little to show, while Zipline treats the work as selling teleportation.
The company and its leader must scale fast. Keller says the pace makes it hard to keep up, and he relies on people who tell him when he is wrong. Zipline avoids middle management and hires hungry, low-ego, practical problem solvers.
Tech needs a clearer optimistic vision. Keller says the industry has failed to present optimistic values, which helps explain public hostility to AI data centres. He predicts abundance for 100bn people and an earth that feels like a nature preserve.
Safety and real-world operations
Three large real-world test sites expose the aircraft to the strange problems of real conditions. Nesting animals and extreme temperatures both cause trouble, and solar weather disrupts GPS, which dual band RTK GPS counters.
The aviation safety bar rises from one problem in 10,000 flights to one in 100,000, then one in a million and finally one in 10m.
Tap a name to go there · tap a line to see why they are linked
On this episode 2
Show 1
Zipline began with blood transfusions and now serves 5,000 hospitals and health facilities, saving 17,000 lives a year. Customers include Walmart and Chipotle.
Backyard delivery with the aircraft about 100 meters up is safer than sidewalk scooters, and Keller argues cities should give the earth back to humans.