This dataset comprises flight logs from simulation tests from the paper: Drone-based Autonomous Motion Planning System for Outdoor Environments under Object Detection Uncertainty.

The UAV sysem was illustrated with an emulated Search and Rescue mission, where a person last seen in a bushland needs to be found.

The probabilistic-based motion planner was tested with the victim placed in two locations. The first one in an open area, and the second surrounded by obstacles (trees). Three flight modes were assesed to compare the performance of our apporach with a baseline motion planner. Each flight mode and vicitm location was tested for 20 iterations.

This record contains the following items:

• Flight logs from PX4 autopilot.
• ROSbags with topics such as processed frame, victim detection status, local position of the UAV and TFs.
• KMLs of recorded GNSS position coordinates of the victim per flight mode.
• Auxiliary scripts to replay the recordings in bash and Python (requires a local installation of  ROS melodic to replay the ROSbags).