FCCfirstclass
To me it is amazing as there is no GPS underground and so the directions are given by lasers.
GPS has nowhere even close to the accuracy available with light. Even a run of the mill total station theodolite has an accuracy of a few millimeters over distances greater than a mile. Laser interferometry is astonishingly accurate. The system used to detect the first gravity waves was able to measure changes in distance of a fraction of the diameter of a proton. The limiting factor is accurately determining the exact location of the starting point, but once this is known (in this case it only needs to be known relative to the ending point) the second point can be very accurately determined indeed.
Typical GPS accuracy is about 3 meters. This can be enhanced by averaging many readings over a period of time (the little flying saucer on a tripod that you see near highway construction is an antenna that allows this), and by integration with WWS signals where they are available to get within a few millimeters in the best case, but the signal can be distorted by nearby objects and atmospheric conditions. And each GPS point is individually located with an inherent error relative to any previous point.