LIGHTNING

Our robot is designed to solve the Rescue Maze task through a combination of mechanical stability, reliable sensing, and optimised software. Its structure was developed from the ground up by our team, with every major component designed to fit the physical limits of the competition while remaining durable and easy to maintain. A key mechanical feature is the kinematically balanced rear suspension, which improves contact with the field on ramps, speed bumps, and uneven tiles, while the low and central centre of mass helps the robot turn without tipping. For victim handling, we developed a precise rescue kit dispensing system that supports accurate rescue kit placement.

The robot navigates primarily with distance sensors and uses a wall-based map representation to plan movement, avoid dead ends, and return to the start. Victim detection is performed with OpenMV cameras using colour filtering, shape analysis, and separate logic for cognitive targets and Greek-letter victims. The main software is organised into a memory-conscious state-based architecture, allowing movement, mapping, victim handling, and restart recovery to work together reliably. What sets our robot apart is the tight integration of custom mechanical design, sensor placement, camera processing, and maze-solving logic, developed and tested step by step through continuous testing and improvement.