Shop Cleaning Robot Navigation

The Shop Cleaning Vacuum project was tasked with creating a autonomous shop cleaning vacuum for Professor Don Heer’s personal shop. As the navigation team, we created the control system that navigates the robot around the shop to clean and back to the charger. To control our robot we used ROS melodic, a robot control framework , running on a Nvidia Jetson Nano and a wide angle camera. Using the camera, we tracked the location of QR codes that labeled the charger and restricted areas. Then by utilizing the hector simultaneous localization and mapping algorithm we processed depth data into a two-dimensional map of the environment and path. Using the described technologies we were able to create a navigation system to control the shop cleaning vacuum. The largest challenge that we faced was working on the navigation of a robot without access to the robot for testing. This was due to other senior project teams developing the chassis and the motor control system at the same time. However, we were still able to develop and test our system in isolation using remote development techniques such as virtual linux machines. In the future, the system could be improved by implementing intelligent pathing for efficient cleaning.

0 Lifts 

Artifacts

Name Description
Project Closeout Document outlining our timeline, requirements, future recommendations, and achievements for our project   Download
Executive Summary Summary of our project.   Download
Project Summary Video describing our project, including main achievements and takeaways.   Link