Project Goal:

Build and program a three-wheeled UGV which utilizes lidar-based SLAM to autonomously navigate a maze. UGV must be able to locate crates in the maze and move them from point A to point B.

System Architecture:

• PID control loops for wheel speed, forward body speed, and rotational body speed

• Lidar-based SLAM for mapping and localization

• IMU + wheel speed sensors for odometry

• A* path planning

• OpenCV-based computer vision for crate location and distance estimation

Hardware:

• Nvidia Jetson Nano processor

• Raspberry pi pico microcontroller

• Lidar unit

• 3d-printed forklift for crate handling

Performance:

The vehicle demonstrated exceptional navigational and control stability while mapping and exploring the maze:

1. Clear map creation

2. Good crate recognition and tracking ability

3. A* demonstrated good functionality in path planning through the maze