The economical interest of robots for construction sites is not yet fully asserted. Existing experimental robots are still complex and costly machines. Most of these future robots are likely to be mobile robots. It is then essential to research simple technical solutions to ensure the mobility function. Attempts to use guiding devices developed for industrial purposes have given informations on the technical and economical limits of many systems. The use of the environment structuration of a building under construction to guide a mobile robot is probably a suitable way to develop simple navigation systems. This paper reviews some of the existing systems, presents an original experiment and draws out the lines of future research that will probably lead to propose modifications of the construction techniques in order to be able to use simple and cheap navigation systems.