A lot of those movements are there to zero out the axes so that each movement starts from a known good position and orients itself against the camera. Usually there's a switch that, for example, senses when the body goes all the way to the bottom, which is the origin for the whole positioning system. Several other movements are for safety since it doesn't have a bunch of cameras and really complex logic for collision avoidance so it resets to a smaller profile between moving around.

Since motors are capable of very precise movements and errors accumulate, this is a best practice when starting new movements. Humans instead have a complex hand-eye coordination system that trained all of our lives (and some people are better at it than others).