544_lab3

Prerequistes

Please install these prerequistes:

pip3 install opencv-python
pip3 install numpy

How to build

• clone the repo such that the folder interfaces/ and lab3/ are located inside your src/ directory of your ros2 workspace
  • e.g your directory might look like:

  ros2_ws/
    build/
    install/
    launch/
    src/
      interfaces/
      lab3/
  

• cd into your workspace root (e.g. ros2_ws/) and then the run the commands in order:
  • colcon build --packages-select interfaces
  • colcon build --packages-select lab3
  • . install/setup.bash

How to run

Simulation

1. Follow the lab instructions to open Gazebo and RViz for the OGM

2. ros2 run lab3 controller will create the controller node. This can be activated all time since it only continuously updates the current pose of the robot until a service request is put it.

3. ros2 run lab3 client <start_x> <start_y> <end_x> <end_y> will start the client node, which will send the start and end position as a Point request to the controller. The controller will perform A* on these start and end points on a pre-generated OGM, and run a controller on each node to reach the target destination. Client ends when the target has been reached. If the Gazebo has just been loaded, a sample client command that works would be: ros2 run lab3 client 1 -0.9 3 3

For the first run, the controller node will generate the map and the metadata as ScaledMap.png and scaledmap.json. These can be loaded in by the controller to run the controller without relying on RViz. After you ran the above steps, to run in offline map mode:

1. In lab3/lab3/kp_controller.py ln 16, change to OFFLINE_MAP=True
2. Ensure that on the same file, ln 61 and 72, it points to the correct path of the metadata and ogm image
3. Re run colcon build --packages-select lab3 from the ros2 workspace