Abstract The technology/designs that are used for manufacturing in engineering industries in developing countries, such as India, are usually bought from developed countries. Since the quality of raw materials, skills of workmen and the environmental control are usually poorer in developing countries the variability in the components is higher than the design requirements. Carrying out 100% inspection and sorting of incoming lots are not practicable owing to high cost. Again rejection of too many lots leads to the unavailability of components on the shop floor and deteriorates the relationship with suppliers. Under this circumstance it is often required to make a compromise by setting wider specifications (called working specifications), which allows incidents of ‘tolerance stack-up’ during assembly or production of non-conforming final products. These, in turn, result in loss of productivity and quality-related problems. This case study carried out in an Indian company demonstrates that the ‘tolerance stack-up’ and quality related problems can be reduced considerably under the existing constraints by defining appropriately (i) the working specifications for the multi-dimensions of a mating component and (ii) the sampling plan for monitoring its incoming quality. This led to a reduction of incidents of ‘tolerance stack-up’ from 19.2% to 3.9%, and usage of the optimal sequence of ordering of the dimensions for inspection minimized the inspection cost.
[1]
A. Raouf,et al.
An optimal sequence in multicharacteristics inspection
,
1990
.
[2]
D. Montgomery,et al.
GAUGE CAPABILITY ANALYSIS AND DESIGNED EXPERIMENTS. PART II: EXPERIMENTAL DESIGN MODELS AND VARIANCE COMPONENT ESTIMATION
,
1993
.
[3]
David H. Baillie,et al.
Multivariate Acceptance Sampling—Some Applications to Defence Procurement
,
1987
.
[4]
Douglas C. Montgomery,et al.
Introduction to Statistical Quality Control
,
1986
.
[5]
Abraham Golub,et al.
Designing Single-Sampling Inspection Plans When the Sample Size is Fixed
,
1953
.
[6]
W. Edwards Deming,et al.
Out of the Crisis
,
1982
.
[7]
David C. Hamilton,et al.
A comparison of methods for univariate and multivariate acceptance sampling by variables
,
1995
.