Multiattribute acceptance sampling plans - Library(ISI Kolkata ...
Multiattribute acceptance sampling plans - Library(ISI Kolkata ...
Multiattribute acceptance sampling plans - Library(ISI Kolkata ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Part 0 : Introduction<br />
0.1 Purpose of <strong>sampling</strong> inspection<br />
The very purpose of quality management comprising of quality planning, quality control,<br />
quality assurance and quality improvement is to achieve simultaneously, improvement in<br />
customer satisfaction, improvement in productivity and reduction in costs of product at all<br />
stages. From the very beginning statistical methods have been developed, established and<br />
implemented as support activities for the above purpose. While Shewhart (1931) focused his<br />
attention on economic control of quality of manufactured product, Dodge and Romig (1929)<br />
attended to quality assurance activities of an organization for protection of the interest of<br />
the internal and external customers at reduced cost of inspection and cost of production,<br />
and also for attainment of uniform quality. The latter established methodologies of <strong>sampling</strong><br />
inspection for separating lots or batches of mass produced items into lots of satisfactory and<br />
unsatisfactory quality.<br />
0.2 Theory and application of <strong>sampling</strong> inspection - a short review<br />
In the following sections we try to underline the nature of major developments of techniques<br />
of <strong>sampling</strong> inspection for industrial quality assurance since the pre-war time. In this context<br />
we limit our discussions at the beginning to single <strong>sampling</strong> <strong>plans</strong> as employed for lot-by-lot<br />
inspection by attribute. The developments relating to double and multiple <strong>sampling</strong> <strong>plans</strong><br />
for attributes, continuous <strong>sampling</strong> and chain <strong>sampling</strong> <strong>plans</strong>, sequential <strong>sampling</strong> <strong>plans</strong> and<br />
their modification and the whole subject matter of <strong>sampling</strong> <strong>plans</strong> by variables and all methods<br />
pertaining to bulk materials are scrupulously excluded from our discussions as they are<br />
not relevant as far as the scope of the present thesis is concerned. The historical perspective<br />
discussed focuses attention only to that part of the development which is conceptually and<br />
theoretically related to the problems studied in the present thesis. A single <strong>sampling</strong> plan<br />
is defined by means of three parameters c, n, and N and the following rule : From each lot<br />
of size N take a random sample of size n. If the number of defectives in the sample is less<br />
than or equal to the <strong>acceptance</strong> number c, accept the lot, otherwise reject the lot.<br />
2