Design and Testing of A Sampling Plan for Two Continuous Production Lines

This paper presents a plan CSP-1-2L for inspection of two continuous production lines. The plan is defined by 2 numbers i₁ and i₂ (the numbers of consecutive non-defective units that must be produced on lines 1 and 2 during a 100% inspection of the lines) and 2 fractions f₁ and f₂ (the specified sampling frequencies for lines 1 and 2 during a fractional inspection of the lines). CSP-1-2L computes 3 performance measures, average total fraction inspected, average total outgoing quality and average total outgoing quality limit, for given values of incoming fractions of defective units on line 1 (p₁) and on line 2 (p₂).

            The validity and accuracy of the performance measure formulas for four inspection patterns have been tested by extensive simulations. Optimal inspection patterns have been found for the sets of p₁, p₂, i₁, i₂, r₁(=1/f₂) values for which the formulas are valid. For low levels of incoming fractions of defectives p₁ and p₂, the formulas for all performance measures are valid for all inspection patterns. However, for higher levels of p₁, p₂, or for large differences between them, the formulas are valid only for a restricted range of i₁, i₂, r₁, r₂ values. Further, if p₁ < p₂ then we must use i₁ ≤ i₂ and r₁ < r₂, and if there is a large difference between p₁ and p₂, then there must also be large differences between i₁ and i₂ and between r₁ and r₂.

Keywords: Continuous sampling plan, Continuous production lines.

Corresponding author: E-mail: g4585017@ku.ac.th and fengpsa@ku.ac.th