For this class, our task was to develop an autonomous navigation system for a differential drive robot (nicknamed "TurtleBot"). The robot would use EKF SLAM for environment mapping + localization, and would use A* for trajectory mapping to a goal location.
The robot had two modes. The first was to reach a goal pose set in RViz, navigating around the environment/obstacles to get there using A*.
The second was autonomous Frontier Exploration, where the robot would continuously explore its environment, until it recognized a stop sign. At this point, it would stop for 3 seconds, before continuing along its path until it saw another stop sign.
Gazebo + RViz: Path Planning & Obstacle Avoidance
Here's a Gazebo demo of trajectory planning + optimization.
Here, we can see the robot plans its trajectory around obstacles.
Interactive Demo: A* vs. RRT*
Click to interact with the demo.
