ROBNEWS
That phrase refers to the handoff between a restaurant kitchen and a customer's door, where DoorDash says its marketplace had no control.
DoorDash spent years as a marketplace that rented other people's drones. On Wednesday, the company unveiled a vertical integration play: a purpose-built aircraft, ground infrastructure, and merchant-handoff software it developed in-house under the DoorDash Air banner. The reason, DoorDash says, is a problem the company could not solve as a broker, the first and last ten feet of a delivery, where food leaves a kitchen and meets a customer's door.
For most of the past several years, DoorDash's drone deliveries were powered by partners. Wing and Flytrex flew the orders; DoorDash supplied the marketplace, the demand, and the app. That structure worked for the easy middle of a delivery, the part that happens a few hundred feet in the air, where the drone can cruise and a GPS coordinate will do. It did not work as well at the seams, where an order sits in a parking lot waiting for a handoff, a restaurant staffer breaks flow to package and load a stranger's aircraft, or a customer walks out to meet the drone. Those are the ten feet.
The aircraft DoorDash unveiled on Wednesday has six slow-spinning propellers, a winch that lowers and raises the package so the drone does not have to land, and a set of failsafes the company says handle weather, range, and kitchen prep variation. The pilot begins in Northern California, with Chipotle, Popeyes, and Momo N Curry as the first named partners. The full stack ties into the existing Autonomous Delivery Platform, which decides whether a given order is a drone candidate based on weather, flight range, traffic, order characteristics, and how long the kitchen actually needs to make the food.
DoorDash estimates under-five-minute restaurant-to-door delivery times in initial tests, but a footnote on the company's announcement limits that claim to a sample of U.S. Marketplace deliveries of limited distance and size and explicitly says the times are not guaranteed. DoorDash also says about 80% of today's typical restaurant orders are light and small enough to fit the aircraft, with a footnote restricting that estimate to U.S. restaurant deliveries of limited distance, size, and measurement period. Payload compatibility is not the same as actual flight eligibility.
The earlier DoorDash Air announcement explicitly described Wing and Flytrex as complementary partners, not ones being replaced. DoorDash has completed what the company describes as hundreds of thousands of lifetime autonomous deliveries across Dot and its robot and drone partners collectively, a cumulative figure that includes the new in-house aircraft only at the margins. The shift is build-and-keep-buying, not displacement.
Unit economics for autonomous delivery are still unsettled across the industry, and DoorDash did not publish cost-per-delivery numbers on Wednesday. The Federal Aviation Administration's beyond-visual-line-of-sight (BVLOS) framework still constrains where and how drones can fly, and DoorDash's Part 135 status is company-reported, not independently confirmed in the announcement. Merchant handoff burden, meaning how much extra work a Chipotle or Popeyes staffer takes on to load a drone, is an open operational question. The pilot's data, when it lands, will be the test.
DoorDash is now running both partnerships and an in-house stack, using the latter to control the part of the delivery closest to a customer's door. Industry coverage of the announcement called the shift DoorDash going "all-in" on its own ecosystem. The company's own language is more measured: a build decision on the ten feet, with partners kept in the loop above and below it, and the long-term test being whether the in-house stack beats the partnerships on cost and reliability at scale.
The next test is the Northern California pilot data, particularly the under-five-minute claim's sample limits and any unit-economics disclosure that follows.