Akademik Veri Yönetim Sistemi
Robotics and Autonomous Systems, cilt.200, 2026 (SCI-Expanded, Scopus)
Mini and micro unmanned aerial vehicles (mUAVs), which are often used as “eyes-in-the-sky” in defense applications are often vulnerable to sniper attacks coming from the ground. In this study, we propose an add-on defense module for mUAVs, inspired by the defensive strategy of honey bees, which upon being hit by a projectile, can be used to estimate the approximate location of the attacker. Specifically, the module estimates the external force vector from the IMU readings recorded within the short time window after the impact, which can be combined with GPS and altimeter readings to predict the attacker's location. The module, designed as an independent add-on for the mUAV, can then broadcast the location, to allow other mUAVs to avoid the attacker. Towards this end, this study proposes, implements, and evaluates both deterministic and Kalman-based probabilistic methods to estimate the external force at the time of the impact. The performance of the methods is systematically evaluated in both simulated scenarios, as well as in laboratory settings using a motion capture system as a source of ground truth. In laboratory settings, the methods estimated the external force detection with an average deviation of 1 degree in the 2D. However, the accuracy of predicting the location decreased along the attack direction due to the effect of gravitational acceleration on the IMU data. The study also highlights limitations, such as the effect of gravitational acceleration on estimation accuracy and the potential impact of other sources of error. Overall, this study provides promising results and offers avenues for future research to improve the accuracy of disturbance force estimation algorithms for predicting the direction and location of ground attacks.