The improvements of robust optical 3D metrology and robot systems are opening new possibilities of inline optical geometry measurements. The combination of these two facts allows measurements using the robot as manipulation system to increase the measurement area beyond a single field of view. By examining the sample from different angles using the robot to obtain high-resolution 3D datasets also from undercuts, measurements which have to date not been possible due to light/shade influences can now be achieved. This method is important on big samples with complex geometries. Special algorithm uses the position and angle of the robot tool centre point, in this case the focus point of the measuring system, for the pre-alignment of the different 3D datasets before the final stitching is performed. Further algorithm is used to improve the overall accuracy through dataset alignment. In this paper, measurement examples are presented, which examine and aim to minimize the uncertainty of this optical 3D metrology and robot combination and are tested against international standards.