Competition Problem: How do we navigate a course autonomously?

First and foremost IGVC is a autonomous competition, the purpose of this page will to be to explain, in a high-level what our approach for autonomous navigation in the competition is. Theoretically, any future members of the team should be able to read this and understand how to approach this problem even if the entire repository is wiped. That is my goal for this passage. I will link to specifics of how we plan to fix certain problems or address certain steps of that autonomous navigation.

The IGVC (Intelligent Ground Vehicle Competition) competition consists of obstacles, guides and waypoints https://en.wikipedia.org/wiki/Intelligent_Ground_Vehicle_Competition. In the third part of the competition (which will be the focuse on for the purposes of narrowing the scope of this wiki) each team must have their robot navigate flat grassy terrain, between white lines in the grass at all times, avoiding cones sawhorses, and other obstacles on its way to several gps way-points through out the course.

In order to approach the problem of autonomous navigation, we've decomposed the problem into 3 parts, Detecting Obstacles, Creating a History of the Environment, and Navigating known Terrain. If your robot can't detect obsticals, it can't navigate around them. If your robot doesn't create a history of the environment, it can't backtrack efficiently or possibly ever find its way out of a maze (it might get stuck in a loop). With out navigation in a known environment, you can't navigate around what you've found