ROBOTICS ENGINEERING

Set up and verify a working robotics development environment before Lesson 1 begins, so class time starts on code, not troubleshooting.

• Linux environment setup (Ubuntu 24.04 via WSL2 or VirtualBox)
• ROS 2 (Jazzy Jalisco) toolchain installation
• Gazebo installation
• VS Code setup with C++, Python, CMake, and ROS extensions

Assignment: The Health Check
A short Python script that tests your ROS 2 environment. Submit a screenshot of the output to unlock Lesson 1.

Get oriented with the course roadmap, meet your instructor, and explore how autonomous mobile robots differ from traditional automated systems.

• Course roadmap and structure
• Assignments and final project overview
• Case study: AMRs vs. traditional AGVs in factory settings
• Robotics challenges: control, navigation, and perception
• Safety challenges across road, indoor, air, and water environments

Learn the ROS 2 computation graph and build your first packages using Colcon workspaces.

• ROS vs. ROS 2 and the DDS middleware layer
• Nodes, topics, services, actions, and parameters
• ROS 2 workspaces and the Colcon build system
• Package structure and dependencies
• Turtlesim node walkthrough
• rqt and rqt_graph introspection tools

Develop Python publisher and subscriber nodes to share continuous data streams between robotic components.

• Topics and the publisher/subscriber pattern
• ROS 2 CLI topic tools
• Workshop: Inspecting turtlesim_node and teleop_key

Assignment #1: Number Publisher
Modify a number publisher node to subscribe to a 'jump' topic and continuously publish the updated cumulative value.

Construct a 3D structural model of a differential drive robot using URDF and visualise its coordinate frames in RViz2.

• Coordinate frames and the TF tree
• URDF tags and robot description files
• Xacro reusable components
• robot_state_publisher and joint_state_publisher
• Demo: URDF construction and RViz2 visualisation

Assignment #2: Add a Shape
Add a shape or volume on top of the robot's URDF model.

Write modular Python launch files to spawn your robot's digital twin into a Gazebo physics simulation.

• ROS 2 launch file structure and rules
• robot_state_publisher, joint_state_publisher, and RViz2 launch
• Gazebo simulator setup and the ROS-Gazebo bridge
• Turtle3 Waffle robot description and CAD import

Assignment #3: Drop the Robot In
Modify the launch file to successfully spawn the Turtle3 Waffle robot into Gazebo.

Implement parameter-driven ROS 2 controllers to actuate robot joints and manage wheel velocities in simulation.

• Motors vs. actuators and wheel encoders
• ros2_control framework and controller plugins
• ROS parameters and YAML configuration
• Workshop: Building a joint velocity controller package

Apply TF2 broadcasters to track and calculate a robot's pose transformations in 2D and 3D space.

• Robot pose representation and transformation vectors
• TF2 static and dynamic broadcasters
• Euler and quaternion angle conversion
• rosbag recording basics
• Workshop: Exploring the TF tree with TF2 tools and rqt graphs

Program differential drive kinematics to translate overall robot motion commands into wheel speeds and estimate position through odometry.

• Forward and inverse kinematics
• cmd_vel and the Twist messaging interface
• Demo: Deriving and implementing differential kinematic equations
• Local vs. global localisation
• Odometry estimation from wheel encoders

Integrate and simulate IMU and LiDAR sensors in Gazebo to capture real-time environmental and orientation data.

• Case study: Sensor working principles and limitations, IMU and LiDAR
• Demo: IMU and LiDAR simulation and integration
• Demo: PlotJuggler for visualising IMU output
• Gaussian noise and probability distributions
• Bayes' theorem fundamentals

Assignment #4: Camera Integration
Integrate a camera sensor into the robot simulation.

Implement a basic Kalman filter to merge noisy odometry and IMU data for accurate robot pose estimation.

• Case study: How self-driving cars and AMRs fuse LiDAR and camera data
• Case study: Measuring real sensor noise with a time-of-flight sensor
• Combining Gaussian distributions
• Kalman filter implementation in Python

Assignment #5: Noise and Behaviour
Change the system's noise value and hypothesise the causation and effect on robot behaviour.

Manage ROS 2 Lifecycle nodes and Quality of Service profiles to safely handle mapping states and long-running navigation processes.

• Global localisation and occupancy grid maps
• ROS 2 Lifecycle nodes and state management
• Demo: Lifecycle node states via CLI
• Quality of Service: reliability and durability settings
• Nav2 framework overview

Configure Nav2's Adaptive Monte Carlo Localisation to dynamically track a robot's position within a known map.

• Markov and Monte Carlo localisation
• Particle filters and resampling
• Adaptive Monte Carlo Localisation (AMCL) configuration
• Demo: Global localisation with good vs. poor initial guesses

Assignment #6: Mapping (SLAM)
Drive the robot through an unknown simulated factory to scan the environment and save a 2D floorplan.

Generate a 2D floorplan of an unknown environment by deploying the Nav2 SLAM Toolbox to map and track the robot simultaneously.

• Particle filter SLAM, EKF SLAM, and Graph SLAM
• Nav2 slam_toolbox configuration
• Demo: Mapping error from wheel radius inaccuracies
• Loop closure

Configure the Nav2 stack to calculate optimal, collision-free paths using search algorithms like Dijkstra and A*.

• Graph search: breadth-first, depth-first, Dijkstra, A*
• Path planning: PID, dynamic window, model predictive control, pure pursuit
• Workshop: Configuring a Nav2 Dijkstra planner
• Case study: iRobot Roomba random bounce vs. Neato XV-11 LiDAR navigation

Assignment #7: Path Planner
Configure an A* path planner and manage its execution using Nav2.

Implement layered cost maps to let the robot dynamically detect and navigate around moving obstacles in real time.

• Layered cost maps and Nav2 costmap 2D
• twist_mux and twist_relay
• Camera integration via the ROS-OpenCV bridge

Synthesise perception, navigation, and control modules into a master state machine that executes a fully autonomous patrol mission.

• System integration and edge case handling
• rosbag debugging and rqt tools
• Workshop: Connecting perception, navigation, and state machine nodes
• Dynamic obstacle recovery

Assignment #8 (Capstone): The Autonomous Inspection Agent
Write the master state machine script connecting all modules so the robot autonomously wakes, patrols, logs anomalies, and returns to its dock.

Present your finalised autonomous robot project and gain actionable strategies for building a robotics portfolio and passing industry interviews.

• Resume tips
• GitHub portfolio structuring
• The robotics interview process
• Capstone project showcase