Inside Zipline's Autonomous System: 140M Miles, Zero Incidents

Keller Rinaudo Cliffton said Zipline’s customers care about outcomes—fast, reliable delivery—rather than whether it’s “a drone.” He framed the company as building an automated logistics system for Earth that can feel like “teleportation,” with the aircraft being only a small portion of what makes that experience work.

Keller Rinaudo Cliffton described early Rwanda operations as a “total disaster,” with Zipline serving only one of 21 contracted hospitals for nine months. He said the team learned that the vehicle itself is only about 15% of the system, while the rest is software, operations, maintenance, inventory, and integration with aviation and healthcare systems.

Eric Watson recounted doctors and lab staff asking why Zipline was only open 12 hours a day since “people get sick 24/7,” which he took as a sign of product-market fit. Keller Rinaudo Cliffton said they moved to 24/7 within the first year and now operate 24/7/365 as dependence grew across thousands of facilities.

Eric Watson explained Zipline designs for rare faults by assuming any component can fail and building redundancy, including two flight computers plus an arbiter that selects which one is in charge. He cited a recent real-world event where the system switched to the backup computer after a delivery and flew home safely, which he linked to achieving 140 million miles with no safety incidents.

Eric Watson said real-world operations introduced unexpected challenges like solar flares that degrade GPS signals via ionospheric effects. He described designing GNSS/navigation to remain robust and adding redundancy beyond GNSS so operations can remain safe even when conditions get “really bad.”