In light of the apparent inability of alternative strategies to guide robots satisfactorily, in real-world settings of daily operation, relative to work pieces without precise, permanent fixture as required by "teach/repeat" , the present dissertation proves real-world robustness and workability of the presented paradigm with actual tasks found in industry. Two fundamental experiments -- thickness-reduction-gauging and five-component positioning -- prove experimentally that the high precision, robustness, and workspace-extent versatility of CSM can be extended to general robot operations. Three types of general applications are implemented based on the present method. The first, peg-in-hole assembly, is a challenging robotics application in industry today, one that requires high precision and robustness of five-component positioning. The second, surface-finishing, task requires a combination of gauging the surface change and positioning the tool with high precision. The third, palletizing and de-palletizing, applications demonstrate the robustness of three-dimensional image analysis in real-world tasks.