What Is an AQL Sampling Plan and How Do You Read the ISO 2859-1 Table?
If you inspect incoming parts or release finished goods, you have probably seen a quality inspector pull a small number of units from a large batch and make a pass/fail call. That decision is not guesswork—it follows an internationally accepted statistical sampling system. The most widely used framework is the Acceptance Quality Limit (AQL) sampling plan defined in ISO 2859-1:1999 (identical to GB/T 2828.1-2012 and historically derived from MIL-STD-105E). This article explains what it is, how to read the tables, and how to apply a single sampling plan correctly.
What It Is
An AQL sampling plan is a set of rules for deciding whether to accept or reject a lot based on inspecting a random sample. The key number is the AQL—the worst tolerable process average percentage of nonconforming items. An AQL of 1.0, for example, means that a supplier's process averaging 1% defectives is considered acceptable in the long run.
The system uses three inputs:
From these, the tables give you a sample size code letter, the sample size (n), and the acceptance and rejection numbers (Ac and Re).
How It Works: Steps to Use the ISO 2859-1 Tables
The procedure is standardized and requires no statistics knowledge—just careful table reading.
- If the number of nonconforming units ≤ Ac, accept the lot.
- If the number ≥ Re, reject the lot.
A Worked Illustrative Example
Example data (illustrative only) — the table values cited are from the published ISO 2859-1 / GB/T 2828.1 standard tables.
Scenario: You receive a lot of 2,000 electronic components. The contract specifies General Inspection Level II and AQL 1.0.
If you had found 3 nonconforming units, you would reject the lot because 3 ≥ Re (3).
Common Pitfalls
Get the Numbers Right Without Manual Table Lookup
Reading the ISO 2859-1 tables is straightforward but easy to get wrong when you juggle multiple lot sizes and AQLs. To avoid errors and save time, use the free Sampling Assistant tool at https://www.6sq.com/tools/sampling/ — it instantly returns the correct code letter, sample size, and Ac/Re pair for any lot size and AQL you enter.
What It Is
An AQL sampling plan is a set of rules for deciding whether to accept or reject a lot based on inspecting a random sample. The key number is the AQL—the worst tolerable process average percentage of nonconforming items. An AQL of 1.0, for example, means that a supplier's process averaging 1% defectives is considered acceptable in the long run.
The system uses three inputs:
- Lot size (how many units are in the batch)
- Inspection level (usually General Inspection Level II, the default)
- AQL value (chosen by contract or specification)
From these, the tables give you a sample size code letter, the sample size (n), and the acceptance and rejection numbers (Ac and Re).
How It Works: Steps to Use the ISO 2859-1 Tables
The procedure is standardized and requires no statistics knowledge—just careful table reading.
- Determine the lot size (N). Count the exact number of units in the batch to be inspected.
- Choose the inspection level. Unless specified otherwise, use General Inspection Level II.
- Find the code letter. In Table 1 (Sample Size Code Letters), locate the row for your lot size range and the column for your inspection level. The intersection gives a letter from A to R.
- Select the AQL value. This comes from your quality agreement, e.g., AQL 1.0 or AQL 2.5.
- Read the single sampling plan. Go to Table 2-A (Single Sampling Plans for Normal Inspection). Find your code letter row and your AQL column. The table shows two numbers: the sample size (n) and, below it, the Ac/Re pair (acceptance number / rejection number).
- Inspect the sample. Randomly draw n units from the lot and count the nonconforming units.
- Make the decision.
- If the number of nonconforming units ≤ Ac, accept the lot.
- If the number ≥ Re, reject the lot.
A Worked Illustrative Example
Example data (illustrative only) — the table values cited are from the published ISO 2859-1 / GB/T 2828.1 standard tables.
Scenario: You receive a lot of 2,000 electronic components. The contract specifies General Inspection Level II and AQL 1.0.
- Lot size: N = 2,000.
- Inspection level: II (default).
- Code letter: In Table 1, lot size 1,201–3,200 at Level II gives code letter K.
- AQL: 1.0.
- Plan from Table 2-A: For code letter K, the sample size is n = 125. Under the AQL 1.0 column, the acceptance/rejection numbers are Ac = 2, Re = 3.
- Inspection: You randomly select 125 units and find 1 nonconforming unit.
- Decision: Since 1 ≤ Ac (2), accept the lot.
If you had found 3 nonconforming units, you would reject the lot because 3 ≥ Re (3).
Common Pitfalls
- Using the wrong inspection level. Level II is the default; Level I reduces sample size (less protection), Level III increases it (more discrimination). Do not switch without a documented reason.
- Confusing AQL with the lot percent defective. AQL is a process average, not a per-lot limit. A lot can be accepted even if its sample defect rate slightly exceeds the AQL—the Ac/Re numbers already account for sampling risk.
- Reading the arrow rules. In the AQL table, if an arrow appears instead of Ac/Re, follow the arrow to the first code letter that has a value, and use that code letter's sample size—not the original one.
- Mixing normal and tightened inspection. The example above is for normal inspection. After repeated rejections, the standard requires switching to tightened inspection (Table 2-B), which has smaller Ac values for the same AQL.
Get the Numbers Right Without Manual Table Lookup
Reading the ISO 2859-1 tables is straightforward but easy to get wrong when you juggle multiple lot sizes and AQLs. To avoid errors and save time, use the free Sampling Assistant tool at https://www.6sq.com/tools/sampling/ — it instantly returns the correct code letter, sample size, and Ac/Re pair for any lot size and AQL you enter.
No related results found
Invited:
6SQ Tools
0 replies