euROBIN Nancy Competition Flight Dataset

This flight dataset provided by the GRVC Robotics Lab of the University of Seville is derived from the euROBIN Nancy Robotics Competition (25-28 November 2024), as part of the paper "Fully Autonomous Dual Arm Aerial Delivery Robot for Intralogistics: the euROBIN Nancy Competition Flight Dataset" submitted to the IROS 2025 conference, whose authors are Alejandro Suarez, Jorge Pozas-Guerra, and Anibal Ollero.

The dataset comprises 42 folders sorted by date and time, corresponding to 54 flight tests. Note that some log files contain two flight tests. These log files were generated by Ardupilot, the open source flight control software [1][2] implemented on the CUAV v5 autopilot. A MATLAB script called DataViewer.m is provided on each folder, so users can easily visualize the data, including the time evolution of the multi-rotor position, velocity, attitude, angular rate, and differential PWM in pitch, as well as the 3D trajectory represented along with the 3D map of the scenario. The differential PWM corresponds to the difference between the duty cycle applied to the Electronic Speed Controllers (expressed in microseconds) of the front propellers with respect to the rear propellers. This differential signal is useful to identify when the arms move forward or when the parcel was grasped as it corresponds to the moment/torque in pitch exerted by the aerial platform controller to compensate the variation in the center of mass.

The 3D map is a point cloud (.PCD file) generated with FAST-LIO [3] using a Livox Mid-360 LiDAR, rotated 30 degrees downwards in pitch. The map was generated moving the aerial robot around the scenario by hand (not flying) before conducting the flight tests. The map included in this dataset (scanThursdayDroneArea.pcd) was generated on Thursday 28 November, corresponding to the scenario of the final demonstration. However, the position of the supply points from which the aerial robot grasped the parcels remained fixed during the four days of the competition. Only the delivery points changed between scenarios. A video from the aerial delivery operation performed by the robot can be seen in [4]. The aerial robot is an improved version of the platform employed for the euROBIN 1st Robotics Hackathon [5], achieving fully autonomous operation in the aerial parcel grasping and drop.

A detailed description of the euROBIN Nancy Robotics Competition can be found on the rulebook [6]. The aerial robot was evaluated in four variations of the scenario, with increasing level of complexity: 

 > 25 November: obstacle-free scenario
 > 26 November: scenario with virtual obstacle
 > 27 November: scenario with physical and virtual obstacle
 > 28 November: scenario with two physical obstacles and one virtual obstacle

As said before, the two supply points were the same in the four scenarios, changing the location of the boxes used as drop points according to the location of the obstacles.

CREDITS

The GRVC Team Members participating in the euROBIN Nancy Robotics Competition were: Alejandro Suarez, Jorge Pozas-Guerra, Jose Antonio Hernandez, and Anibal Ollero.

CONTACT

Alejandro Suarez (asuarezfm@us.es)
GRVC Robotics Lab, University of Seville
https://grvc.us.es/

ACKNOWLEDGMENT

This work is supported by euROBIN, the European ROBotics and AI Network (Grant agreement 101070596), funded by the European Commission.

REFERENCES

[1] https://ardupilot.org/
[2] https://ardupilot.org/copter/docs/common-downloading-and-analyzing-data-logs-in-mission-planner.html
[3] https://github.com/hku-mars/FAST_LIO
[4] https://www.youtube.com/watch?v=Yg_AyYJrI28
[5] A. Suarez et al., "Door-to-Door Parcel Delivery From Supply Point to User’s Home With Heterogeneous Robot Team: The euROBIN First-Year Robotics Hackathon," in IEEE Robotics & Automation Magazine, doi: 10.1109/MRA.2024.3501954.
[6] https://www.eurobin-project.eu/images/2024/euROBIN_Nancy_Coopetition_RuleBook.pdf