Economic optimisation of acceptance interval in conformity assessment for processes with known and unknown biases

The optimisation of the acceptance interval in the conformity assessment of total inspections for processes with known and unknown biases is discussed from the perspective of maximising the producer’s revenue. It is assumed that the unknown bias and the random measurement errors in the inspection are derived from normal distributions with known variances. To maximise the revenue function, we propose a highly efficient algorithm. Moreover, comparisons with existing methods for determining the acceptance interval are conducted. We conclude that the proposed approach is superior because: (i) the economic model used is based on the concept of global consumer’s risk, modified to express the actual risk borne by consumers, (ii) both known and unknown biases, the latter of which has a significant impact on optimisation, are explicitly addressed, and (iii) maximising revenue provides an incentive for producers to set a guard band for the acceptance interval. • The optimisation of the acceptance interval for conformity assessment is proposed for processes with known and unknown biases, and inspections involving measurement uncertainty. • A practically fast numerical solution for this optimisation has been developed using Monte Carlo simulation with importance sampling. • Comparative analysis with other techniques reveals that our proposed method is superior because it can explicitly handle unknown biases, which are often significant in practical applications.