AI will make drone threats a nightmare – it could also save us

Paul Lemmo - Lockheed Martin 

Source Link

Imagine you’re standing watch on a military base’s security forces team, monitoring for potential air threats. Your radar screen is cluttered with tracks – commercial aircraft, flocks of birds, and civilian and commercial drones. You see what looks like a small aircraft veering towards your fence line.

Is that a threat? A delivery drone flying off course? A hobbyist drone operator snapping a picture of a sunset? Or is it an AI-trained distraction while the real threat masks its approach from another direction?

Drones already present a formidable challenge for military and homeland security forces. AI-guided drones could be a nightmare, using advanced tactics to hide their movements or coordinate swarming attacks custom-designed to overwhelm defenses.

The good news? AI is uniquely suited to powering exceptional drone defenses. Here’s how:

Learning algorithms are excellent at spotting and tracking drones: In a noisy or cluttered radar environment, drones could slip through cracks in sensor detection. But AI can be trained specifically to separate signal from noise in a given environment.

An AI-powered system can become an expert in the area around a military base, for example, learning the local landscape, structures, and even weather patterns so it knows how to pick out and track drone anomalies with exceptional accuracy.

AI can match defensive weapons to drone targets much faster than humans: Once the drone is spotted, a counter-UAS system has to know how to best determine intent and to plan for mitigation. But that depends on many complex factors. Is the drone carrying explosives? Is it vulnerable to a cyber or electronic attack? Could a laser take it down safely?

For operators in a command center, making those decisions from a dozen or so football fields away can cost precious minutes. But an AI algorithm can be trained to recognize and evaluate different drone threats in an instant, evaluating its weaknesses and capabilities to quickly find and recommend the best way to maintain safety and sovereignty. It can also be trained on policy and rules of engagement to know which responses best align to regulations, assisting operators in a complex data rich environment.