Most of infrastructures for civil engineering (highway road, bridge, tunnel, etc.) built around the urban region in high-growth era of 1970-1990 begun to reach the life, and it has been the time when large-scale repair and renewal of infrastructures became necessary. Especially, with the collapse accident of the tunnel generated some time ago in Japan, the inspection of the superannuated tunnel was had requested. Usually, in the periodic inspection of the road tunnel, maintenance engineers approached the wall by using mobile elevating work platform as much as possible, and then the deteriorated wall parts were detected by sighting and hammering sound. To progress the automated inspection and maintenance of inner wall of tunnel, the advanced inspection system with restricting the traffic regulation was developed. In this inspection system, the guide frame along tunnel wall was installed on the protection unit stepped over the road like gantry crane. The inspection device moved with stability by adopting the guide frame, and the inspection accuracy could be improved. However, when this unit moved along the tunnel, this guide frame should avoid the convex obstacles such as duct fan, lamp and several road traffic sign in the tunnel. Therefore, by composing the entire frame of VGT (Variable geometry Truss), the shape of guide frame was changed flexibly and it could be passed in the tunnel. As a shape control of the guide frame, the inverse analysis method was applied such as robot manipulator combining two fixed-length members in series. The angle of each frame was reversely analyzed according to the shape of the obstacle measured with the laser sensor, and the actuator of the frame was controlled simultaneously. Applying such control method, the shape was transported smoothly to keep a steady structure of guide frame. In this paper, the outline of tunnel wall inspection applied variable guide frame, the structure principle of the guide frame, basic method of shape analysis and control technology, the finding of convex obstacles by laser sensor and the whole inspection system are explained in detail.