Drone Navigation Using Cameras and Echolocation

This 2017 patent describes a drone navigation system that uses both cameras and echolocation sensors to detect obstacles and find its way, with echolocation taking priority for immediate threats.

What it covers

This patent details a navigation system for aerial vehicles, like drones. It combines a GPS transceiver for general location, an optical system (cameras) providing a wide 180-degree view to spot visible obstacles, and an acoustic system (echolocation) with a narrower 10-degree view to detect obstacles, especially smaller ones not visible to cameras. A central processor uses all this data to plot a course to a destination. If the acoustic system detects an obstacle directly in the path, it overrides the GPS and camera data to command an immediate dodging maneuver. Once the obstacle is clear, the drone returns to its original path. The optical system can even provide a full 360-degree view, and the acoustic system can cover up to 45 degrees.

What it doesn't cover

• —Navigation systems that only use cameras and no echolocation sensors.
• —Navigation systems that only use echolocation sensors and no cameras.
• —Navigation systems where GPS data always overrides sensor data for obstacle avoidance.
• —Navigation systems that do not have a processor to calculate dodging maneuvers.
• —Systems for ground-based vehicles; this patent is specifically for aerial vehicles.

The clever bit

The core innovation is how the system prioritizes immediate, short-range acoustic sensor data over GPS and camera data when an obstacle is detected. This ensures rapid reaction to close-proximity threats, which is crucial for preventing collisions in dynamic environments.

Why it matters

This patent addresses a key challenge in autonomous flight: reliable obstacle detection and avoidance in diverse conditions. By combining multiple sensor types with a prioritized processing logic, it enables drones to navigate more safely and effectively, even when encountering unexpected or hard-to-see objects.

Real-world examples

• 1.Aurora Flight Sciences drones
• 2.Modern autonomous drone navigation systems
• 3.Advanced aerial vehicle guidance systems