Viewed README.md:1-216

Based on the project's documentation (README.md), you ran the ROS 2 simulation using a multi-terminal setup. You used 4 to 5 separate terminals to launch the different components of the robot.

For each terminal, you always start by sourcing your workspace:

source /opt/ros/jazzy/setup.bash
cd ~/ros2_ws
source install/setup.bash

Here is exactly how you ran each part of the project:

Terminal 1: Launch Gazebo Simulator

You started the 3D world with the furniture/objects and spawned the robot:

ros2 launch diff_drive_robot spawn_robot.launch.py world:=yolo_world.sdf

Terminal 2: Start Navigation & SLAM

After waiting about 5 seconds for Gazebo to load, you launched the SLAM toolbox and Nav2 stack to handle mapping and localization:

ros2 launch diff_drive_robot navigation.launch.py use_sim_time:=True use_slam:=True

Terminal 3: Run the Semantic Brain Node

This is the main Python node you have been working on (semantic_navigator.py). It handles the YOLO object detection, PID movement, and logic:

ros2 run diff_drive_robot semantic_navigator

Terminal 4: Start Manual Teleop Control

You used this custom script to drive the robot around with your arrow keys during the "scanning" phase:

ros2 run diff_drive_robot arrow_teleop

Terminal 5: Voice Commands (Optional)

If you were testing the offline voice control (Vosk), you ran this node to speak commands like "go to chair 1" instead of publishing them manually:

ros2 run diff_drive_robot voice_commander

---

How you commanded it:

If you weren't using voice commands, you told the robot what to do by publishing string messages to its command topic from any terminal:

• Start scanning: ros2 topic pub --once /semantic_nav/command std_msgs/String "data: 'scan'"
• Stop scanning: ros2 topic pub --once /semantic_nav/command std_msgs/String "data: 'scan stop'"
• List objects: ros2 topic pub --once /semantic_nav/command std_msgs/String "data: 'list'"
• Go to object: ros2 topic pub --once /semantic_nav/command std_msgs/String "data: 'chair_5'"