pic: N Dimensional Exploring Tree

[Marlow]

To expand on @faust1706's explaination:

My method constructs 2 RRTs, one from the initial state, the other from the goal state. The program is constructs these 2 trees independently and simultaneously. After every few iterations, the program checks to see if the shate a point, or roughly share a point, in space. It goes until they do, then the path is going down the tree based at the initial starting point and then it goes up the goal tree from the shared point.

While this method isn't the greatest for low dimensionality search spaces, it shines when the dimensions start getting big as it scales linearly with dimensions.

I'm curious to see if any team will ever actually implement this. Alas I don't think teams have the ability to given the amount of moving pieces it has. It seems like it'd be a nightmare.