Description of running all the nodes in the integration branch

Overview

This project consists of several nodes that work together to control a robot. The nodes are managed using a tmux session, which allows for easy parallel execution and monitoring of multiple processes. The start_all_nodes.py script is used to initialise and start all the necessary nodes for the robot's operation.

Starting All Nodes

To start all the nodes, execute the start_all_nodes.py script. This script sets up a tmux session with a 3x3 grid of panes, each running a specific node of the project. The script will automatically kill any existing session with the same name to ensure a fresh start.

Steps to Start All Nodes

1. Ensuretmux is installed on your system. You can install it using your package manager, e.g., sudo apt-get install tmux on Ubuntu. But it should already be installed by the setup scripts

2. Run the script:

   Bash

   python3 nodes/start_all_nodes.py

3. Monitor the nodes: The script will create a tmux session named dev-environment. You can attach to this session to monitor the nodes:

   Bash

   tmux attach -t dev-environment

Detach from the session: To detach from the tmux session, press Ctrl + B followed by D.

1. Reattach to the session: If you need to reattach to the session later, use the attach command mentioned above.

2. To kill the session: To kill the tmux session, press Ctrl + B followed by typing :kill-session and hit enter.

Node Descriptions

node_imu.py

• Purpose: This node handles the Inertial Measurement Unit (IMU) data. It initializes the IMU, calibrates it, and publishes both raw and processed IMU data to the MQTT broker.
• Key Features:
  • Calibrates the IMU on the first cycle.
  • Publishes gyro and raw IMU data at a specified rate.
  • Publishes to the TOPIC_GYRO, TOPIC_RAW_IMU and TOPIC_RAW_IMU_UNSCALED_BIASED topics

node_control.py

• Purpose: This node is responsible for controlling the robot's motors based on the target velocity and path plan. It uses a lookahead controller for trajectory planning.
• Key Features:
  • Subscribes to target velocity and path plan messages.
  • Publishes wheel velocities to the MQTT broker.
  • Generates a stopping velocity profile when no target velocity is available.
  • Publishes to the TOPIC_WHEEL_VELOCITIES topic

node_localization.py

• Purpose: This node processes IMU and wheel encoder data to estimate the robot's pose using an Extended Kalman Filter (EKF).
• Key Features:
  • Initializes with static IMU data.
  • Publishes processed IMU and localization messages.
  • Corrects EKF state with encoder data.
  • Publishes to the TOPIC_EXTENDED_POSE_W_BIAS, TOPIC_PROCESSED_IMU, TOPIC_ROBOT_POSE, TOPIC_ROBOT_EXTENDED_POSE. TOPIC_LOCALIZATION_INITIALIZED topics

node_mapping.py

• Purpose: This node manages the depth wavemap and generates occupancy grids for mapping the environment.
• Key Features:
  • Integrates depth images to update the wavemap.
  • Publishes occupancy and traversability grids.
  • Publishes robot pose grid coordinates.
  • Publishes to the TOPIC_OCCUPANCY_GRID, TOPIC_TRAVERSABILITY_GRID, TOPIC_ROBOT_POSE_GRID_COORDS, TOPIC_WAVEMAP_OCCUPIED_POINTS topics

node_rerun.py

• Purpose: This node visualizes the robot's data using the Rerun dashboard. It processes input data and logs it for visualization.
• Key Features:
  • Subscribes to various MQTT topics for data visualization.
  • Logs occupancy grid, TOF map, odometry, and path plan data.

node_drivepath.py

• Purpose: This node handles path planning using the D* algorithm. It generates a path from the robot's current position to a target point to be consumed by the control node.
• Key Features:
  • Subscribes to target point and traversability grid messages.
  • Publishes path plan messages.
  • Replans the path periodically or when the target point changes.
  • Publishes to the TOPIC_PATH_PLAN topic

Web Interface

The project includes a web interface for controlling the robot and visualizing its environment. The interface can be accessed through a web server running on the robot.

Accessing the Web Interface

1. Start the web server: The web server is started as part of the start_all_nodes.py script. It serves the web_interface.html file.

2. Open the web interface: Access the interface by navigating to the following URL in your web browser, replacing the color with whatever your bracketbot name is:

   http://color-bracketbot.local:8080/web_interface.html

3. Features:

   • Control the robot using on-screen buttons or keyboard keys.
   • Visualize the robot's environment and pose on a grid.