Our position-correcting router relies on a field of high contrast dominoes - which are easy for the computer vision system to identify - in order to locate itself.