Hands-On ROS for Robotics Programming

Scenario 2 – travel to a target location avoiding the obstacles

The procedure to start the training is similar to scenario 1:

1. We only need to change the name of the files and use the relevant ones:

T1 $ roslaunch gopigo3_gazebo gopigo3_stage_2.launchT2 $ conda activate tensorflowT2 (tensorflow) $ roslaunch gopigo3_dqn gopigo3_dqn_stage_2.launch

2. If you want to see graphically how the GoPiGo3 learning process is performing, execute the result_graph.launch file:

T3 $ roslaunch gopigo3_dqn result_graph.launch

The following screenshot shows all the content you should see on the screen of your laptop:

The red graph shows the Total reward obtained ...