RoboGames 2024 University Category Competition 

 Technical Solution

• Implemented a flood-fill algorithm for efficient pathfinding in the maze environment

• Optimized robot control parameters for smooth navigation and accurate positioning

• Created a solution robust enough to work from any starting position within the maze

Elimination Round

The second round significantly increased in complexity, requiring navigation through a 5m × 5m maze with hazardous areas (fire pits with varying damage zones) and locating three survivors (green squares) scattered throughout the environment.

Technical Solution:

• Implemented two advanced pathfinding algorithms to map the maze:

  1. Trémaux Search Algorithm - An efficient maze-solving technique that marks visited paths to avoid redundant exploration

  2. Depth-First Search (DFS) - Used for comprehensive maze exploration and mapping

• Optimized return-to-start navigation after completing rescue objectives

Finals

In the final round of the competition, we transitioned from simulation to working with physical Kobuki robots in a real-world environment. This stage required adapting previously developed algorithms while dealing with the challenges of hardware limitations, sensor noise, and environmental unpredictability.

Technical Solution

• Developed and deployed OpenCV-based computer vision algorithms to detect and differentiate colored walls under varying lighting conditions. This included color space conversion (BGR → HSV), masking, contour analysis, and real-time filtering for robust detection.
  
  

• Integrated sensor data from wheel encoders, IMU, and IR sensors to improve localization and implement real-time obstacle avoidance.
  
  

• Developed decision-making routines that allowed the robot to autonomously perceive, plan, and act in real-time using onboard vision and sensor data.
  
  

Our efforts culminated in a successful performance in the finals, where we were awarded the 2nd Runner-Up position in the University Category. This achievement highlights our ability to bridge the gap between simulated and physical robotics systems, applying theoretical knowledge to practical engineering challenges.

This final round provided valuable experience in:

• Embedded system integration

• Vision-based robotics using OpenCV

• Autonomous navigation in real-world environments

• Translating simulation logic to physical robots
  
  

The RoboGames 2024 competition was an incredible opportunity to deepen my understanding of autonomous systems and solidify my interest in robotics and intelligent systems development.