Abstract:
The objective of this master project is to compare the efficiency among single (S), double (D), and multiple (M) sampling plans in the PCB production process in order to determine an efficient sampling plan which can reduce customer's complaints occurred from circuit damage problem and can also reduce inspection time. The author proposes three sampling plans based on Operating Ratio method using Acceptance Quality limit (AQL)! Producer's Risk, and Customer's Risk as 0.65%, 0.05%, and 0.10% respectively. Based on this method, the S plan is (55:1), the D plan is (33, 33: 0, 1), and the M plan is (10, 10, 10, 10, 10, 10, 10: *, *, *, 0, 0, 1, 2). For comparison purpose, three indices namely, defects percentage, number of defects per million, and inspection time required, are used. These plans are then implemented for 3 months and results are compared to the current sampling plan. It obviously shows that the current defects percentage is reduced dramatically from 0.01% to 0.00264%, 0.00282%, and 0.00250%, the current number of defects per million is reduced from 176 ppm to 68 ppm, 72 ppm, and 67 ppm, and the inspection time for each plan are 1.8, 2.2, and 1.66 hours (all results sequenced as S,D, and M plans). Based on these results, the M plan is recommended since it obtains acceptable indices and it has more chances to sampling than other plan.