Dataset · 2025
DronePropA: Motion trajectories dataset for defective drones
Data in Brief
An open dataset of 130 indoor flights of a small quadcopter flown with healthy and damaged propellers, so that researchers can develop and test methods that spot a damaged propeller from flight data alone.
Why it matters
A cracked or chipped propeller rarely makes a drone fall out of the sky at once. The drone keeps flying, but the damage shows up as small changes in how it moves and how hard its motors work. Catching those changes early means fixing a drone before it fails, which matters for drones used in delivery, agriculture and public safety. Until now, few public datasets covered enough damage types and severities to train and compare detection methods.
What is in the dataset
- 130 flight logs from three drones of the same model (Quanser QDrone 2).
- Five flight paths: diagonal cross, square, stepwise climb and descent, straight climb, and yawing.
- Three kinds of propeller damage (edge cut, crack, surface cut), each at three severity levels, plus healthy reference flights, at two flight speeds.
- Position, velocity, attitude, thrust commands, battery level and inertial-sensor signals, all recorded at 1 kHz.
How it was recorded
All flights took place in the IRC-ASE flight lab at KFUPM, with a 14-camera motion-capture system giving millimetre-level ground truth. The tests are indoors, so they contain no wind or turbulence.
What it can be used for
- Training classifiers that tell a healthy drone from a damaged one.
- Identifying quadrotor dynamics, where damage appears as a change in model parameters.
- Developing and testing fault-tolerant control.
The data are freely available on Mendeley Data under a CC BY 4.0 licence.
Cite this work
Mohamed A. A. Ismail, Mohssen E. Elshaar, Ayman M. Abdallah, Quan Quan (2025). DronePropA: Motion trajectories dataset for defective drones. Data in Brief. https://doi.org/10.1016/j.dib.2025.111589
@article{ismail2025dronepropa,
title = {DronePropA: Motion trajectories dataset for defective drones},
author = {Mohamed A. A. Ismail and Mohssen E. Elshaar and Ayman M. Abdallah and Quan Quan},
journal = {Data in Brief},
year = {2025},
doi = {10.1016/j.dib.2025.111589},
}