Video Demo

Project Summary

Project objective

Program control logic using readings from LiDAR module to steer a robot through a maze without touching the walls

Summary

Developed a python program to interact with the TurtleBot Waffle's sensors and motors using ROS2. The development cycle began with getting comfortable with ROS2's communication and custom interface packages. Then we began simulating the maze problem in Gazebo and developed the control logic. After a successful simulation maze run, we moved on to the real robot.

Testing our program with the real robot resulted in everything breaking. The lidar module had a completely different protocol than the simulation, the readings were littered with error, the robot's movements were different, and worst of all, it just kept slamming into everything. I worked tirelessly on reading the LiDAR data correctly and filtering out all the error, while my teammates made the motion control more robust.

Eventually, our robot successfully traversed the maze without running into any walls!

Takeaways

• Robotics is a lot of trial and error

• When developing control logic, make it easily adjustable 

• ROS2 makes interacting with sensors and actuators extremely simple

• Simulations are useful, but not always accurate

Slide Deck