Final Project Website
Peter Liu, Josh Nagle, Erick Hillebrand
Our group chose to work on the LADAR project. Our goal was to create a program that would allow the robot create a real-time map of its surroundings using the data read from the LADAR. In order to minimize the computational load on the OMAP, we decided to do as much of the mapping as possible on the desktop computer itself. The robot uses an updated right-wall follow algorithm for navigation and collision avoidance. It uses a combination of IR sensors and LADAR data to achieve this. While it navigates the course, it continually takes readings from the LADAR for mapping purposes.
The LADAR mapping data is sent by Linux over the WiFi connection to the VB mapping program running on the desktop. The VB application reads and saves this data, and updates a map to show the current geometry of the course. The program draws a single point to represent every reading from the LADAR. After readings have been taken for some time, the user can press the “Clean” button. This removes the individual points corresponding to every LADAR reading and replaces them with a cleaner approximation of where the obstacle or wall actually is. This will also remove ghost readings and “footprints” made by a person walking on the course as it is being mapped.
If the course geometry is significantly changed, or if the user wants to restart the mapping from scratch, there is a “Clear” button which clears all the saved data and completely clears the map. The robot will then have to start drawing a new map from scratch.