SE423 - Final Project (Rex the Cyborg Dog)

Project Expectations:

Building upon concepts and code from our lab sessions, this project challenges us to develop an autonomous ball-sorting robot ("Rex") that integrates and expands our previous work with LIDAR navigation, PID control systems, and sensor integration. We'll enhance the basic object collection system from Lab 7 with advanced color detection and sorting capabilities, while improving the path planning algorithms we've experimented with in earlier labs. The project requires us to combine these lab-tested components into a more sophisticated system capable of autonomous operation and obstacle avoidance. While leveraging existing lab code for motor control and basic functionality, we'll implement new vision processing, sensor fusion, and decision-making logic to create a complete working system. The final will be a live demonstration showing Rex autonomously navigating, collecting, sorting, and delivering balls while avoiding obstacles.

About Rex:

Rex is an autonomous robotic system that combines LIDAR-based wall following and dead reckoning for precise navigation. The robot uses LIDAR to correct accumulated dead-reckoning errors, ensuring accurate positioning as it moves through hallways. For object interaction, Rex employs an OpenMV camera with LAB color thresholding to identify and classify colored balls, which are then collected using a servo-operated gripper for sorting and delivery.

Our Strategies:

Hybrid Navigation:

• Dead reckoning for initial path execution
• LIDAR sensor feedback to correct positional drift
• Wall-following logic for structured environments

Vision & Manipulation:

• OpenMV camera with LAB color space filtering for reliable ball detection
• Custom thresholding to classify colors under varying lighting
• Servo-controlled gripper to secure, sort, hold, and release balls

Obstacle Awareness:

• Real-time LIDAR scans to avoid static and dynamic obstacles
• Path re-planning when deviations occur

Built on lab-proven concepts but extended with new sensor fusion and control logic, Rex demonstrates howlocalization corrections (LIDAR + dead reckoning) and adaptive vision systems (OpenMV) can create robust autonomous behavior.

Rex in Action:

Rex being a BAD dog!:

Along the way we had to do a lot of troubleshooting and refinement. Here are some of Rex's best bloopers.

Our Team:

Jacob McBride (Mechanical Engineering '25)

Garrett Sigrist (Mechanical Engineering '25)

Omar Alqazaq (Systems Engineering and Design '25)

Andre Baltazar (Systems Engineering and Design '25)

Cassie Hays (Systems Engineering and Design '25)

Acknowledgements:

Thank you for looking at our project. We could not have done this without the guidance from our amazing TA's as well as Lab director, Dan Block. This class has been real...SENIORS OUT!