The system monitors all drone parts, spots anomalies and enables self-recovery, addressing surging cyberthreats where hackers can hijack unmanned aircraft, causing erratic behavior like course changes, altitude loss or crashes—risks to logistics, infrastructure and rescue missions increasingly reliant on such devices.
Traditional defenses focus mainly on navigation sensors, but hackers can spoof GPS signals or tamper with internal software. SHIELD goes further by analyzing physical traits like battery levels, processor temperatures and power use; even slight fluctuations signal an attack's onset.
“Without robust recovery mechanisms, a drone cannot complete its mission under attacks, because even if it is possible to detect the attacks, the mission often gets terminated as a fail-safe move,” said Mohammad Ashiqur Rahman, lead researcher and associate professor in FIU’s Knight Foundation School of Computing and Information Sciences.
Using machine learning models to identify attack patterns, SHIELD detected threats in 0.21 seconds on average during tests and initiated recovery in 0.36 seconds, halting damage before severe impacts.
Developers liken the method to medical diagnostics: a single symptom may not indicate illness, but holistic physiological scans catch issues early. Similarly, SHIELD examines the drone's "physics" to pinpoint attack types and apply tailored fixes.
As drones proliferate in agriculture, delivery and infrastructure monitoring, demand for such safeguards grows urgent. SHIELD could pave the way for safer skies ahead.