This repository is a companion resource for the Evaluation for Robotics module, ROB2002, at the University of Lincoln. You will find here details about the development environment enabling work with the robots, the description of practical tasks for each week, and practical instructions for the assignment.

Workshops

For a list of all the workshops offered as part of ROB2002 and additional resources, please refer to the Wiki. The rest are instructions on how to use the files and code provided for the evaluation for robotics assessment ROB2002-2425

Robotic Counting (Baseline vs. Enhanced)

This repository contains ROS2 Python nodes and scripts to demonstrate a robotic counting experiment under three different environments:

• Environment 1 (Open warehouse)
• Environment 2 (Moderate obstacles)
• Environment 3 (Heavily cluttered)

9 boxes were placed (3 red, 3 green, 3 blue) in each environment.

Nodes Overview

1. combined_detector.py

   • Subscribes to a color + depth camera feed (topics: /limo/depth_camera_link/image_raw, /limo/depth_camera_link/depth/image_raw, etc.).
   • Detects colored boxes in 2D, converts them to 3D in the global frame (odom), and publishes each detection as a PoseStamped on /object_location.

2. counter_3d.py

   • Subscribes to /object_location.
   • Maintains a list of unique objects, skipping any new pose within 0.3–0.6 m of an existing one.
   • Prints “Total Unique Objects: X” each time a new object is added.

3. baseline_sweeper.py

   • A simple corner-based approach with minimal coverage.
   • Visits corners, returns home.
   • Typically faster but misses occluded objects.

4. enhanced_sweeper.py

   • A multi-waypoint or zigzag approach covering the entire warehouse area.
   • Slower but more comprehensive coverage.

Installation

1. Clone this repo into your ROS2 workspace:

   cd ~/your_ros2_ws/src
   git clone https://github.com/codewitted.git
   cd ..
   colcon build
   source install/setup.bash

Make sure Docker is running on your machine before you do anything else. Go to worlds and run a chosen world. Once it's successfull, in another terminal run the enhanced or baseline node scripts to see the demo. You could also open and launch the maps yaml files to observe the experimemts there.

The results like final “Total Unique Objects” or logs about corners visited go to your console (and can be easily saved to file with standard shell commands, e.g. > output.txt)

Running scripts in the results sections either prints a png of the outcomes or prints the data in terminal.