DAA animation 41
DAA animation 41

Using Sound to Unlock Instant Logistics at Scale

Zipline’s Autonomous Avoidance System

Zipline is expanding rapidly around the world, building the first instant logistics system to serve all people equally. To enable a future where everyone has access to the products they need when they need them, Zipline must be able to fly safely in even the most complex and varied airspace around the world.  This requires us to solve a fundamental technical challenge: how to detect and avoid other aircraft in the air, wherever they’re flying.

When Zipline set out to build a detect and avoid solution, we approached it by talking with our customers and drawing on the data and expertise we have accumulated from years of national scale operations. Other solutions out there failed to meet our standards for serving customers now: Ground-based systems are cumbersome and don’t work in all geographies. Camera-based systems are costly and can’t solve for common visual constraints like clouds.  And systems that require new regulations or industry consensus, like universal traffic management, are years away from real use.

After years of research, development, and testing, we turned to an approach that’s been used in aviation for years - sound. We’ve created a first-of-its-kind onboard microphone based detection and avoidance system. It is small, light, and ready to be deployed at scale, anywhere our customers need us to be, without new regulations or ground infrastructure.

Acoustic Detection

Zipline’s acoustic detect and avoid system bypasses solutions that only perceive in one direction. Instead it relies on a series of small, lightweight microphones that, unlike human pilots, provide 360-degree awareness of air traffic with a range up to 1.25 miles (2 km), through common visual obstructions like clouds.

Intelligent Avoidance

In real-time, a Zipline drone can autonomously and continuously monitor for airspace traffic, identify the intruder’s potential flight paths, and adapt the drone’s own paths to avoid potential hazards and safely complete the mission.

Development and Testing

We put our detect and avoid solution through extensive flight test campaigns to push the system to its limits. Intruder aircraft flew many different representative patterns, with variations in altitudes and speeds, to simulate what our aircraft face in commercial operations. We also tested the system across geographies and weather patterns ranging from North Dakota winters to Ghanaian summers and the Rwanda rainy season. This enabled us to understand and account for the ways terrain and atmospheric conditions cause the sound to travel and reflect differently.

Additional Resources

Zipline's CTO shares the full story

Read more about how Zipline solved autonomy's detect and avoid challenge.