Novel Family of Group Chain Acceptance Sampling Plans for Truncated Life Tests Using Minimum Angle Method

Acceptance sampling is a fundamental statistical quality control procedure widely used in industries to decide whether or not to accept an item lot. When the quality feature concerns lifetime of an item which is too long to be inspected, the testing time is often terminated after a specified duration, and this is termed as truncated life test. Since the decision of lot acceptance is based on the result of sample inspection, there exist risks of (i) rejecting good lots (producer’s risk) and (ii) accepting bad lots (consumer’s risk). Chain sampling has been introduced to overcome the drawback of single sampling which appreciably decreases the probability of lot acceptance even with a small increase in probability of failure, p. However, the chain sampling plans currently used in industry often consider only the consumer’s risk and sequential item testing, although multiple items can be monitored simultaneously in groups. To close these gaps, this study proposes several group chain acceptance sampling plans which enable multiple items testing and consider both producer’s and consumer’s risks simultaneously. By using minimum angle method, both risks are minimized through design of the smallest angle resembling the ideal operating characteristic curve. This angle is calculated by using the tangent function at different values of p. The major benefits are the ability to fairly protect both producers and consumers and achieving more accurate decision using minimum sample size. Additionally, the group chain sampling plans (multiple items testing) will consequently reduce test time, energy, labour and operational cost. Practically, they should offer better alternatives to industrial practitioners in any sectors involving items life test