Precision Landing (Assignment)

The purpose of this mini project is to familiarize yourself with the EOLab System and how ROS2 interplay with the PX4 ecosystem.

In reality, a full-precision landing assignment requires knowledge and skills in computer vision for marker detection, search algorithms, path planning, pattern generation to survey for the marker, and drone controllability. But in order to minimize friction, this project focuses solely on the basics of “controllability” and has implemented only the minimum components for marker detection, deliberately omitting search and path generation.

Execution Overview

The following diagram shows the expected execution workflow.

System Start-Up

The EOLab system provides a set of predefined worlds, which can be specified during launch using the world argument. For this project, launch the world called aruco:

ros2 launch eolab_bringup start.launch.py world:=aruco

As mentioned in the introduction, a node is provided that handles detection and reports a set of results. To start the detector, you can run:

ros2 run eolab_precision_landing detector --ros-args -r __ns:=/protoflyer

Publishes:

Topic	Message type
/<PROVIDED NAMESPACE>/detected_aruco_markers	visualization_msgs/MarkerArray
/<PROVIDED NAMESPACE>/aruco_debug_image	sensor_msgs/Image

Subscribes:

Topic	Message type
/<PROVIDED NAMESPACE>/camera_info	sensor_msgs/CameraInfo
/<PROVIDED NAMESPACE>/image	sensor_msgs/Image

As aditional information keep in mind that to control the drone you should publish the control command on:

Topic	Message type
/<DRONE NAMESPACE>/fmu/in/manual_control_input	px4_msgs::msg::ManualControlSetpoint