Now day car manufacturing industries are ruined by a robot. Robots are participating different types of tasks, such a tasks welding is one of the major operation, most of the industries are spending more time for inspecting the welding state. For this purpose, we propos welded crack detecting robot. This robot is holding. The magnetic flux system (MFS), MFS has been developed for the detection of the welding cracks on the plate. The MFS signal response from the welding sample which the Cracks varies and not the proper joining sample surface. A data acquisition program was developed to continuously display the Cracks variation of the tested sample. The signal analysis technique such as the power spectrum density of the fourier transform is applied to obtain an optimum parameter to describe the wall thinning of the plate. This technique can be used as a potential tool to detect the local wall thinning of a plate.