Taki & Umemoto's UAV control damps payload swing with no extra sensors

Cable-suspended payload transport is a staple for multirotor UAVs in construction, logistics, and agriculture, but the pendulum-like swing of the load creates periodic disturbances that degrade tracking accuracy and can destabilize the aircraft. Traditional anti-swing methods rely on additional sensors (vision, encoders) or precise pre-calibration of cable length and payload mass—both impractical for real-world operations where cargo changes frequently. Taki and Umemoto's approach removes those requirements entirely, using only the drone's existing IMU and throttle signal.

The core innovation is an extended Kalman filter that treats the unknown pendulum frequency as a hidden state, extracting the periodic swing disturbance from IMU measurements contaminated by the drone's own motion. A subsequent active damping controller computes a correction angle that is added to the attitude control loop, effectively dissipating pendulum energy without ever knowing the payload's physical properties. Flight experiments validated the method across a range of cable lengths and mass variations, showing consistent damping performance. This work, submitted to arXiv on August 19, 2026, could significantly simplify deployment of cargo drones by removing the need for payload characterization or external sensing—making autonomous aerial delivery more robust and practical in dynamic field environments.