Detect and Avoid for Drones: Sensors, Standards and BVLOS Operations

Why detect and avoid matters

When a drone flies within visual line of sight, the remote pilot watches for other aircraft. Beyond visual line of sight (BVLOS) — pipeline inspection, long-range delivery, mapping — that is no longer possible. A detect-and-avoid (DAA) capability must find conflicting traffic and obstacles and act in time, either automatically or by alerting the pilot.

Cooperative and non-cooperative traffic

• Cooperative aircraft broadcast their position, for example through ADS-B or other electronic conspicuity devices. Receiving these signals is relatively cheap and reliable.
• Non-cooperative aircraft — gliders, balloons, some light aircraft and other drones — may transmit nothing. Detecting them needs onboard or ground-based radar, electro-optical and infrared cameras, or acoustic sensors.

The DAA function chain

A DAA system typically works in stages: detect the intruder, track it, evaluate whether a conflict is developing, decide on a manoeuvre and execute it, all within the time available. Sensor fusion and tracking algorithms are central, and so is the latency of the command-and-control link if a human is in the loop.

Regulation and safety assessment

In Europe, BVLOS operations in the specific category are commonly assessed with the SORA methodology, which looks at ground and air risk together. In the United States, BVLOS rules are being developed by the FAA. Industry standards such as those from RTCA and EUROCAE define performance requirements for DAA systems. Across jurisdictions the logic is similar: the required robustness of DAA depends on the airspace and the risk of the operation.

Engineering challenges

Small drones have tight limits on size, weight and power, so sensors must be light and efficient. Detection range must be long enough to allow avoidance at the closing speeds involved. And machine-learning-based vision systems must be validated in a way regulators can trust — a problem very similar to the one automotive engineers face in functional safety for ADAS.

Explore our aerospace, UAM, drones and space programmes to work on these systems.