Ratio of random inspections of distribution boxes

The ratio of distribution boxes to inspect randomly is typically determined using statistical sampling plans, often guided by AQL standards, and depends on batch size, inspection level, and acceptable defect rate.

Determining Sample Size

Random inspection involves selecting a subset of boxes from a larger shipment to assess overall quality. The sample size is influenced by:

• Lot size: Total number of boxes in the shipment.
• Inspection level: ISO 2859-1 defines normal and special inspection levels (e.g., Level I, II, III) to balance inspection cost and protection.
• AQL (Acceptable Quality Level): Maximum percentage of defective items considered acceptable. Common AQLs for major defects range from 1% to 2% depending on product criticality . For example, a common industry practice is to select a number of boxes equal to at least the square root of the total number of boxes for a preliminary random inspection . This ensures a statistically significant sample while keeping inspection costs manageable.

Sampling Methods

• Simple Random Sampling: Each box has an equal chance of being selected. Suitable for smaller shipments or when defects are expected to be evenly distributed .
• Cluster Sampling: Boxes are grouped into clusters (e.g., pallets), and entire clusters are randomly selected. Efficient for large shipments but may miss defects if they are unevenly distributed .
• Random AQL Inspection: Often used when working with multiple suppliers or when a small percentage of defects (e.g., 2–3%) is acceptable. The inspector selects boxes randomly according to the AQL table and inspection level .

Practical Considerations

• Defect clustering: Defects may not be evenly distributed; some cartons may contain multiple defective items. This can affect the reliability of the sample, so inspectors may adjust the sample size or use stratified sampling .
• Critical vs. minor defects: Critical defects may require a higher inspection ratio, while minor defects can tolerate lower sampling rates .
• Cost and time: Random inspection reduces labor and equipment costs compared to 100% inspection, while still providing a reliable estimate of overall quality .

Example

For a shipment of 400 boxes with an AQL of 1% for major defects and using normal inspection Level II, the inspector might select 20–25 boxes randomly for inspection. If defects exceed the acceptance number, the batch may be rejected or subjected to further inspection . In summary, the ratio of boxes to inspect is not fixed but is determined by statistical sampling plans, AQL limits, inspection levels, and the expected distribution of defects. Using these guidelines ensures a balance between inspection efficiency and quality assurance.