Occasionally, product classification as merely good or bad is not enough and variable measurements do not apply. For example, in evaluating the quality of a new automobile, there could be many defects but it would be misleading to classify the entire automobile as unacceptable. The solution in situations like this is another attributes chart, the CU-CHART, which monitors the number of defects per inspection unit. In general, the inspection unit is usually expected to have some defects and we wish to know whether the number of defects is excessive. CU-CHART is also valuable when dimensions or units of measure complicate quality assessments. For example, suppose that a coil of steel is 100 meters long and contains 7 lamination defects. What is the defect rate? It could be 7/100 = 7%. But the defects are small, each perhaps a centimeter in length. There are 10,000 centimeters in the coil so the defect rate becomes 0.07%. We could also compute the square centimeters in the area of the coil and compute yet another defect rate. The only sensible way around this problem with dimensions is to state quality in terms of total number of defects per inspection unit.