A Physics-Based Fault Tolerance Mechanism for UAVs’ Flight Controller

Unmanned Aerial Vehicles (UAVs) are on the rise across a wide range of application domains like shipping and delivery, precise agriculture, geographic mapping, and search and rescue. Thus, ensuring UAVs’ safe operations and reliable integration into civilian airspace is essential. These unmanned vehicles face various potential hazards and threats, such as software or hardware failures (e.g., GPS malfunctions), communication failures, and security attacks (e.g., GPS Spoofing), which can threaten mission completion and safety. Thus, implementing a fault-tolerant mechanism to improve the resilience of UAVs is crucial. This research aims to introduce a fault-tolerance mechanism employing a physics-based model that accurately estimates drone positions in the presence of hazardous conditions, particularly in the presence of GPS faults. The physics model that relies on Newton’s Second Law of Motion, enables real-time and precise estimation of the drone’s position in faulty conditions throughout a mission. Thus, the physics model’s values can replace the erroneous GPS input values. The results obtained through our experiments, conducted using fault-injection techniques in a simulated environment, demonstrate the effectiveness of our physics-based fault-tolerant mechanism, particularly in mitigating GPS-related hazards.