Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO 28592:2017, prepared by ISO/TC 69/SC 5, provides double sampling plans by attributes with minimal sample sizes, indexed by producer’s risk quality (PRQ) and consumer’s risk quality (CRQ). It replaced ISO 28801:2011. These plans are designed for isolated lots or short series of lots where switching rules cannot be applied, making them essential tools for quality control in environments where production runs are not long enough to benefit from the adaptive switching mechanisms of AQL-indexed systems. The plans minimize sample sizes while keeping both producer’s and consumer’s risks within specified nominal values.
The plans follow a simple two-stage structure that is intuitive for inspectors to apply. For nonconforming items: if zero nonconforming items are found in the first random sample, the lot is accepted immediately. If two or more are found, the lot is rejected immediately. If exactly one is found — the critical borderline case — a second smaller sample is drawn. If zero nonconforming items appear in the second sample, the lot is accepted; otherwise, it is rejected. This two-stage approach provides significant average sample size savings compared to single sampling while maintaining operational simplicity.
| First Sample Result | Action | Second Sample (if needed) |
|---|---|---|
| 0 nonconforming | Accept lot | N/A |
| 1 nonconforming | Take second sample | Accept if 0, reject if ≥1 |
| ≥2 nonconforming | Reject lot | N/A |
The plans are indexed by nominal producer’s risk quality (PRQ) and consumer’s risk quality (CRQ), with three combinations of nominal risks: (5 %, 5 %), (5 %, 10 %), and (10 %, 10 %). Up to 17 preferred nominal values of CRQ and PRQ are provided for each risk pair, giving users flexibility to match their specific quality requirements. This indexing approach differs fundamentally from AQL-indexed plans — it provides direct control over both error probabilities rather than relying on switching rules across a series of lots for consumer protection. The nominal producer’s risk α and consumer’s risk β are not exceeded by the actual risks of any plan provided in the tables.
The standard provides both curtailed and uncurtailed inspection options. Under curtailed inspection, testing stops as soon as the acceptance or rejection decision is determined. For example, if the first two items in the first sample are both nonconforming, testing can stop immediately with a rejection decision, rather than completing the full first sample. Curtailed inspection always results in average sample sizes equal to or smaller than uncurtailed inspection, and the standard provides both sets of average sample size curves for planning purposes.
ISO 28592 addresses a critical gap in quality engineering: the need for rigorous risk-controlled sampling when inspecting isolated lots or short production runs. In modern manufacturing with shorter production runs and higher quality levels, AQL-indexed systems designed for long-term process average control may not be appropriate. The double sampling plans in ISO 28592, with their minimal sample sizes for given risk levels, provide a statistically sound alternative that explicitly controls both producer and consumer risks.
The plans are particularly valuable in regulated industries (medical devices, aerospace, pharmaceuticals) where demonstrating explicit control of both producer’s and consumer’s risks is a regulatory requirement. The standard’s provision for both nonconforming items and nonconformities per 100 items extends its applicability across different quality metrics and inspection regimes.