Multiaxial vibration data for blade fault diagnosis in multirotor unmanned aerial vehicles

The complete dataset consists of five Excel files, each corresponding to a specific flight condition of the UAV, including both healthy and faulty cases. These cases involve controlled modifications to the UAV's propeller blades, such as mass removal (to simulate damage) and weight addition using adhesive tape (to induce imbalance). The dataset was collected under repeatable indoor hover flight conditions, using a centrally mounted ADXL335 triaxial accelerometer connected to a NI USB-6008 DAQ system. Each file contains approximately 400,000 rows of raw, high-resolution vibration signals recorded continuously over a 7-minute period. The files are named clearly to reflect the specific fault case they represent: •Healthy •Damaged Bottom Right Blade •Damaged Top Right Blade •Unbalanced Bottom Right Blade •Unbalanced Top Right Blade Each Excel file includes four columns, structured as follows: •Column A: Time (in seconds) •Column B: X-axis acceleration (m/s²) •Column C: Y-axis acceleration (m/s²) •Column D: Z-axis acceleration (m/s²) Manual labeling was used to ensure maximum clarity, with filenames explicitly indicating the fault condition and blade position. All data are recorded in the time domain, suitable for transformation into frequency or time-frequency domains depending on user needs. These characteristics make the dataset especially useful for UAV diagnostics, machine learning classification tasks, and vibration signal analysis.