Enhancing UAV survivability through real-time stall detection and parachute assisted recovery

Abstract

Abstract Ensuring the safety and survivability of unmanned aerial vehicles (UAVs) during stall conditions is critical for minimizing operational risks, financial losses, and system failures. This paper presents a novel stall detection and emergency recovery system that integrates an STM32-based flight controller with a multi-parameter, altitude-aware parachute deployment mecha- nism, distinguishing it from conventional approaches that rely solely on single-threshold pitch/roll triggers. Stall events are detected using gyroscope and accelerometer data from an onboard inertial measurement unit (IMU), combined with angular velocity, vertical acceleration, and motor saturation to accurately identify unrecoverable conditions. Emergency responses are triggered within 1.1 s of stall confirmation. System performance was validated through Mission Planner simulations and. Hardware-in-the-Loop (HIL) testing, demonstrating that parachute deployment at 25 m reduced impact velocity from 22.2 m/s to. 3.2 m/s (95% survival rate), while deployments at 10 m proved largely ineffective (survival rate ∼40%), emphasizing the need for low-altitude impact mitigation strategies below 15 m. Additionally, the system transmits GPS coordinates upon landing to facilitate rapid UAV retrieval. The proposed approach offers improved reliability, multi-parameter detection, and hybrid recovery. compared to existing methods, with direct applicability to delivery, surveillance, and industrial inspection drones. The design has been filed under Indian patent number 202,541,051,308.