# RRT demo

The rapidly-exploring random tree is a general sampling-based technique for exploring a metric space1, but I'm mostly interested in its applications as a motion planning algorithm for robots. I made these simulations of RRTs exploring $$\mathbb{R}^2$$ to help myself get an intuition for how RRTs work. One day, I hope to extend these simulations to $$\mathbb{R}^3$$ and other spaces to see how RRTs do in different topologies.

RRT-ANN is a word I made up to refer to RRTs that use an approximate nearest-neighbor subroutine in place of an exact nearest-neighbor one.

There are some strikingly similar examples in nature (video).

## Random expansion

1 LaValle, Steven M. "Rapidly-Exploring Random Trees: A New Tool for Path Planning." (1998). [back]