This project showcases Dawg, a maze-solving robot built to autonomously localize, navigate, retrieve a payload, and deliver it within five minutes. Using a Monte Carlo particle filter and deterministic pathfinding, Dawg completes most tasks autonomously, with manual control for payload handling. It achieved ~90% localization accuracy and consistent performance, all within a $370.92 budget.
- Self-localize within a static maze using a Monte Carlo particle filter
- Navigate autonomously to a target location using deterministic pathfinding
- Avoid collisions with maze walls and obstacles
- Detect and grip a randomly placed block using an ultrasonic sensor and servo-driven gripper
- Retrieve and deliver the payload from the loading zone to the drop-off zone
- Switch between autonomous and manual control for task-specific execution
- Eject the payload precisely at the destination via a controlled spin
- Operate under a random starting position and block placement
- Complete all tasks within five minutes
- Maintain robust and repeatable performance across trials
- Stay under budget with a total cost of $370.92 (below the $400 limit)