When performing spot checks, QIMA inspectors exclusively use ISO 2859 and its tables. This document, published by the International Organization for Standardization (ISO), is an international standard and has equivalents in all national regulations (ANSI/ASQC Z1.4, NF06-022, BS 6001, DIN 40080).

Acceptable Quality Level (AQL) sample is a method widely used to define a production order sample to determine whether the entire order meets the customer's specifications. Based on the sample data, the customer can make an informed decision on whether to accept or reject the batch. The inspection report will clearly indicate whether your product meets the selected acceptable acceptable quality level.

How to determine the correct sample size and acceptance number?

Our trained account managers will ensure that your selected inspection level and AQL values ​​are the best way matched your needs.

Example: For a hypothetical inspection of 4,000 units, the client selected Normal Level II inspection and an AQL value of 2.5.

In Table A below, the intersection of the appropriate lot size and the overall inspection level indicates the sample size with the letter code L. Then, referring to Table B, we find row L, which indicates the required sample size of 200. To meet an AQL value of 2.5, it is acceptable that No more than 10 units of products from a sample of this size were not tested.

Download sample report

Do you need more detailed information?

Packs of gloves are sometimes marked with special markings. One of them is AQL. What is this indicator and what does it mean? Let's try to understand this issue. Acceptable Quality Level– this is an acceptable level of quality medical gloves, or in other words, the percentage of acceptable defects for the entire batch.

Naturally, production on conveyors has its own errors and it is too expensive to control the quality level of each pair 100%. Therefore, manufacturers came up with a method of selective testing that makes it possible to isolate the component of defective products for the entire batch. If, as a result of testing, the value AQL exceeded the permissible limit, then the entire batch is considered unfit for consumption and must be removed from sale.

In different countries, AQL indicators for acceptable values ​​are different. So, for example, the Russian Federation, according to our requirements, inspection points should have a value of no more than 2.5, and European standards require a more stringent framework and correspond to a value of 1.5.

According to AQL values, the following quality indicators are distinguished:

• 1 – very high
• 1.5 – high
• 2.5 – acceptable
• >3 – invalid

Accordingly, the lower the AQL index, the higher the quality level of medical gloves.

Long-term practice shows that most imported manufacturers claiming AQL 1.5 do not correspond to the declared data and have a much higher percentage of defects. Therefore, the benefit that combines low price and low quality is very doubtful. A defective batch will end up costing the same money, but the supplier’s reputation will be damaged and the medical staff will remain dissatisfied.

/ / / /

What does AQL mean?

When using the services of an independent inspection company, it is important to understand the standard used to draw random samples for a cosmetic inspection.

The most common standard for product inspection is ISO 2859-1 (ANSI/ASQC Z1.4-2003). It uses the AQL (Acceptable Quality Limit) concept.

What does AQL mean? The standard definition of AQL is “the maximum percentage of defects in a lot (or the maximum number of defects per hundred units of product) that, for the purpose of quality acceptance of a lot, can be considered to meet average standards and be acceptable.”

Sample size, based on the AQL tables, will be selected and then checked for defects.

Defects are divided into three categories: minor, major and critical. Although different clients classify differently, typical definitions look like this:

• Minor defect– this is a non-compliance with standards, but something that most likely will not affect the use of the product.
• Significant defect- this is the one that is most likely to lead to the impossibility of using the product for its intended purpose.
• Critical defect is one that is considered dangerous or unsafe.

According to the number of defects found and according to the number of defects allowed (the numbers are given in the AQL tables), your inspection company can advise you accept or refuse from the party.

How to use AQL tables?

AQL tables will help us determine the sample size we need for inspection based on the quantity ordered and your level of inspection rigor. You can choose Levels I, II or III, with Level III being the most stringent test and Level I the least stringent. The standard level that is used by default and by 98% of people is Level II. The choice is up to the client, but this is the recommended level.

To find the required sample size for inspection, you must first refer to the first table and find the total number of products produced on the left. For example, if you produce 8000 pieces, in level II you see the letter L, which in the second table represents a sample size of 200.

At the top, the second table lists defect levels from 0 to 6.5 (we cut out the higher values ​​because they don't apply to consumer product buyers).

You can choose which level to apply to your types of defects: critical, major and minor. Typically, most importers choose standard defect levels: 0/2.5/4.0 , but someone chooses 0/1.5/2.5, this is at the request of the client.

Using standard defect levels of 0/2.5/4.0 and a sample size of 200, we see that if you have more than 0 critical defects, 10 major defects and 14 minor defects, you need to reject that lot.

Of course, the decision on what to do after receiving the inspection results is up to you. Most importers prefer to discuss what was found during the inspection with the supplier/manufacturer to improve anything that can be improved. In case the inspection results are very close to the AQL limits, it is important to double-check whether the level of defects found is acceptable to you or not.

Almost any industrial enterprise that mass-produces its products faces the task of statistical acceptance control.

What is statistical acceptance control? To answer this question, let us turn to the Great Soviet Encyclopedia. So:

Acceptance statistical control – 1) this is a set of statistical methods for monitoring mass products in order to determine its compliance with specified requirements; 2) this is an effective means of ensuring the good quality of mass products.

This paper describes one of the types of statistical acceptance control - control by alternative characteristics.

All procedures for this type of control are standardized and described in GOST R ISO 2859-1-2007 “Part 1. Sampling plans for successive lots based on an acceptable level of quality” And GOST R 50779.72-99 “Part 2. Sampling plans for individual batches based on the maximum quality level LQ”.

The purpose of these procedures is to help influence the supplier economically and psychologically through the possible rejection of lots, as well as to help maintain an average level of process quality, while ensuring that the risk of accepting lots with low quality (consumer risk) is at the proper level.

In addition, these procedures provide

• automatic consumer protection in cases of detection of a decrease in quality,

• incentive to reduce control costs while achieving a stable level of quality.

This paper describes both the theoretical aspects of acceptance control procedures based on an alternative criterion, and a software package that implements these procedures based on a powerful statistical package STATISTICA.

Description of the task

Within the framework of this work, the following task is considered:

A large set of objects of 2 types is considered. It's about the party N products in which n products turn out to be defective (i.e. have at least one defect), and the rest N-n are suitable.

With the unknown n you can estimate this number from a relatively small sample, choosing at random (randomly) m products and identifying all of them as defective (let their number be equal to k).

Let's calculate the probability of the event , where is the random number of defective products in the sample.

The probability of this event is described by a hypergeometric distribution and is equal to

, Where k= 0, 1, …,

Based on the observed number, one can estimate the proportion of defects (input defect level) in the entire batch of volume N.

From a statistical point of view, during acceptance inspection, a hypothesis is tested H 0: that the party is suitable, against the alternative H 1: that the batch is unsuitable.

Here q 0- the value of the percentage of defects when the batch is still considered suitable.

Typically, during sampling inspection, batches are divided into passable and unfit using two numbers - AQL (acceptance defect level) and LQ (rejection defect level). Lots are considered acceptable when and unusable when q>LQ. At AQL< q < LQ (так называемая область неопределенности) качество партии считается ещё допустимым.

Acceptance defect level AQL is the maximum permissible value of defect level in a batch manufactured during normal production. The quality rejection level LQ is the limit for classifying products as defective.

According to GOST procedures, when monitoring on an alternative basis, so-called “sampling control plans” are used.

The control plan establishes the number of units of production from each batch ( m), subject to control and the necessary acceptance criteria for the lot.

Acceptance standards are used as a batch acceptance criterion ( c) and, sometimes, rejection numbers. The batch is accepted if the number of defective units in the sample is m < c .

The specific sampling plan is selected based on the following factors:

1. level of control (normal, enhanced, weakened)

2. batch size

3. AQL level

Based on these factors, a specific sampling plan is selected according to GOST tables, i.e. sample size m and acceptance number c.

Each sampling plan is characterized by its own operational characteristic (OC).

The operational characteristic of the control plan is the function P(q), equal to the probability of accepting a batch with a share of defective units of production q.

, where is the probability of occurrence k defective units of products in a sample volume m.

Most often, the operational characteristic is displayed in the form of a graph, where

, at q= AQL

, at q= LQ.

Here is the supplier's risk, equal to the probability of rejecting a batch with q= AQL (type I error of accepting alternative " H 1: batch rejection" versus hypothesis " H 0: acceptance of the party"); - consumer risk, equal to the probability of accepting a batch with q= LQ (type II error).

By setting the risk levels and , based on a specific sampling plan, the corresponding AQL and LQ numbers can be calculated, on the basis of which a decision is made whether to accept or reject the entire lot.

Calculator

The task described in the previous section was implemented in the form of a probabilistic Statistical Acceptance Control (SAC) calculator.

This version of the calculator implements the calculation of the parameters of a one-stage sampling plan, in accordance with GOST R ISO 2859-1-2007.

This application is written on the Visual Basic .NET platform using statistical package libraries STATISTICA.

The calculator has an intuitive user interface and is easy to use.

The calculator interface is implemented in the form of two windows:

1. welcome window

2. main dialog box

After launching the calculator, the user sees a welcome window.

Rice. 1. Welcome window

While this window is displayed, the system is loading in the background. STATISTICA.

After the system STATISTICA loaded, the welcome window disappears and the main calculator dialog box appears.

Rice. 2. Main dialog box

In the main window you can set the following characteristics of the sampling plan:

After specifying all the plan parameters, you must press the “Calculation” button.

Based on the first three characteristics entered, the calculator will calculate and plot the Operational Characteristics curve for this plan.

In addition, the calculator will mark the risk levels of the supplier and consumer on the OX curve and calculate the corresponding AQL and LQ numbers.

Rice. 3. OX curve

Example

As an example, let us consider the acceptance inspection of bearing batches using an alternative criterion.

Let the batch size of bearings be 150 pieces. According to the procedures of GOST R ISO 2859-1-2007, we must select a sample size code. At the “general - II” control level, a batch volume of 150 units corresponds to code F.

We consider a one-stage sampling plan. We are looking for the required plan in the GOST tables based on the sample size code F and the AQL level of 2.5%.

The following plan corresponds to these parameters:

• sample size N = 20

• acceptance number c = 1

Launch the SPK calculator. Enter all the necessary plan parameters. The default risk levels for supplier and consumer (according to GOST recommendations) are set to 5% and 10%, respectively.

Rice. 4. Example of using a calculator

After clicking the “Calculate” button, the calculator will display the OX curve and calculate the AQL and LQ numbers for the risk levels we have selected.

In this example:

AQL = 1

LQ = 4

Those. To organize acceptance inspection of bearings with a batch size of 150 units using single-stage normal sampling control with supplier risk levels of 5% and consumer risk levels of 10%, it is necessary:

Case – production of fasteners

Consider an enterprise producing fasteners (various bolts, nuts, etc.) for the automotive industry.

Obviously, this type of product is characterized by increased requirements for its reliability and, consequently, acceptance control.

There are numerous specialized standards (both foreign DIN, ISO, and Russian GOST) that fully describe the requirements for parameters, tolerances, etc. fasteners

Let's look at the production of fasteners in more detail.

The technological process for producing bolts consists of the following main stages:

• planting of workpieces;

• transportation;

• thread rolling.

When landing the workpiece, the mechanic controls all dimensions of the bolt:

• bolt length (L);

• diameter of the smooth part (d);

• hex size (AV);

• head height (k).

Rice. 5. Bolt parameters

When rolling threads:

• outer diameter of thread d outer = d;

• average thread diameter d avg ;

• thread length (b);

• thread pitch.

Based on thread length, bolts are divided into 3 types:

1. standard (standard thread length according to the bolt size group);

2. full (thread up to the bolt head);

3. without thread.

Let us describe an example of organizing acceptance control at this enterprise.

Let's say this enterprise has two bolt production lines. Each line, after appropriate adjustment by a mechanic(s), is capable of producing finished bolts with specified parameters (for example, 20 pieces per minute).

Naturally, carrying out complete control in this case is a very costly undertaking (the expenditure of time and resources is enormous). Therefore, one of the optimal options may be single-stage sampling control based on an alternative criterion.

This type of control is distinguished by its simplicity and cost-effectiveness compared to continuous product control.

Organizing a one-stage selective acceptance inspection is indeed very easy.

Let's say we assembled a batch of bolts of the same type, manufactured on one line, set up by a specific team of mechanics, with a volume of 1000 pieces (in approximately one hour of line operation).

Next, you need to randomly select a certain number of bolts from this batch.

Next, these bolts are examined for compliance of their parameters (geometric dimensions) with the standards described in the GOST corresponding to this bolt. If any parameter of the inspected bolt does not meet the standard, then the bolt is considered defective.

Based on the number of defective bolts (or the number of defects), we can estimate the number of defective products in the entire batch.

However, we are faced with the following questions:

1. how many bolts to select for inspection,

2. What is the probability of mistakenly rejecting a good batch, or, conversely, accepting an unsuitable batch?

Naturally, the more bolts we remove from a batch for subsequent inspection, the less likely it is to make a mistake, but the greater the cost of organizing inspection.

To find the optimal sample size value, you can use the Statistical Acceptance Control (SAC) calculator, developed on the platform of a powerful statistical package STATISTICA using GOST standards in the field of acceptance control.

The operating principle of the SPK Calculator is described in detail in the first part of this material.

Having set the level of risks of the supplier and consumer, the volume of the batch, based on the standard level of non-conformities, using the SPK calculator, you can select the optimal sample size, acceptance and rejection numbers.

Let's look at a specific example.

• Batch volume – 1000 pcs;

• Standard level of non-conformities NQL - 0.65%;

• Supplier risk – 5%;

• Consumer risk – 10%.

Based on the GOST tables, we are suggested to use the following plan:

• Sample size – 367 pcs;

• Acceptance number – 1 pc.

Let's enter the described parameters into the calculator and press the calculate button.

Rice. 6. Operational characteristics of the control plan for bolt production

Based on the results obtained, the acceptance number is taken to be 1 piece, the rejection number is 4 pieces.

Those. if for a batch of 1000 bolts out of 367 randomly selected, only one defective bolt was found, or no such bolts were found at all, then such a batch is considered acceptable.

If in the same situation 4 or more defective bolts are found, the batch is rejected (sent for continuous inspection).

In other situations, the batch is subject to additional control (taking another sample or using multi-stage plans).

By collecting statistics on the number of passable/unfit batches for each production line, all teams of mechanics and other production parameters, we can draw conclusions about the possible causes of defects in production.

This information will help the manager outline an action plan to improve the situation with product quality, which in turn will allow:

• increase the competitiveness of goods on the market;

• reduce costs by switching to more cost-effective control plans;

• increase sales volume, and, consequently, the income of the enterprise.

Bibliography

1. Rozanov Yu.A. Probability theory, random processes and mathematical statistics, Nauka, 1985.

2. Ivchenko G.I., Medvedev Yu.I. Mathematical Statistics, Higher School, 1992.

3. Borovikov V.P. Popular introduction to the program STATISTICA, Computer Press 1998.

4. Borovikov V.P., Borovikov I.P. STATISTICA. Statistical analysis and data processing in the Windows environment, Filin 1998.

5. Borovikov V.P. STATISTICA, the art of data analysis on a computer, St. Petersburg 2001.

6. Encyclopedia. Probability and mathematical statistics. (Chief editor Yu.V. Prokhorov), Moscow, Great Russian Encyclopedia, 1999.

7. GOST R ISO 2859-10-2008 “Statistical methods. Alternative sampling procedures. Part 10. Introduction to the standards of the GOST R ISO 2859 series.

8. GOST R ISO 2859-1-2007 “Statistical methods. Alternative sampling procedures. Part 1: Sampling plans for successive lots based on acceptable quality levels.”

9. The activities of specialists are aimed at increasing the efficiency of quality departments of textile factories. Textile Control works with textile products at all stages of production and shipment. The company provides a full range of services, provides consultations on quality control at all stages of production and on finished products, helps in the formation of full-fledged quality control departments in factories for enterprises from Uzbekistan.

   AQL inspection from Textile Control

   Carrying out an AQL inspection based on scientific calculations reduces the time and financial costs of quality control and increases its efficiency.
   

   What is AQL?

   AQL implies an acceptable (guaranteed) level of quality at which the maximum possible number of defects in a batch of goods is allowed relative to the zero mark.

   For the client, the ideal is zero defects in a huge batch of products. But in the realities of mass production, any production process allows for a certain number of errors, and no manufacturer guarantees 100% absence of defects. The compromise is based on strict control of the entire production process - from the stage of supply of raw materials to the shipment of the final product.

   AQL is an effective measure to bring product quality to the required level, guaranteeing consistently high quality.

   How does the international standardized AQL system work?

   The Textile Control team - pioneers in Uzbekistan - introduced the AQL quality control system inspection standard MLT-std 105, providing monthly monitoring of enterprises in Uzbekistan upon customer request. When conducting an AQL inspection, CLIENTS who have ordered a batch of textile products from a particular factory receive objective data on the required parameters from Textile Control specialists. Among them:

   • list of products
   • product quality
   • objective assessment of the supplier’s integrity
   • visualization and remote presentation of the enterprise you are interested in
   • assessment of the actual presence or absence of production capacity and infrastructure at the factory.

   Textile Control specialists carry out all types of inspections covering the full production cycle. Among them: test-sample, at the production stage, final-pre-shipment, quality control of the shipped order, and also accompany the order in the “from scratch and turnkey” format.

   Types of inspections: Initial, In-line, Final Inspection

   Initial Inspection
   When samples first enter the production line, we carry out an incoming inspection. This is a test-sample inspection, during which samples are taken selectively and their quality is monitored based on the measurements of the patterns, their compliance with the cut parts, the coincidence with the colors and print patterns and other parameters. At your request, we send test samples for approval to put them into production. Initial inspection will significantly reduce the percentage of possible defects in the further production process.

   In-line Inspection
   When the samples have entered the production line and the process of sewing products has already started, we carry out an In-line inspection, checking the goods at the production stage, selectively removing samples from the line and performing quality control of tailoring, color and other details before packaging.

   Final Inspection: pre-shipment inspection and monitoring of the loading process
   Finally, the batch is ready for shipment. At the final stage, we conduct a pre-shipment inspection, sampling in a chaotic manner. We test samples in many respects, from measurements to reading bar codes on labeled products and checking the thickness of packaging boxes, based on documentation provided in advance by the customer, and strictly focusing on description and art-work (“digital” and artistic descriptions of the product ). In the process of loading consignments of goods, we carefully check the number of pieces of cargo in the container, the integrity of the packaging and monitor compliance with the pre-agreed loading scheme.

   Each stage is accompanied by constant reports of all stages of production up to the shipment of a consignment of goods - we keep a photo report and broadcast it to the customer in real time, ensuring the effect of the client’s presence at all stages of production.

   Thanks to AQL inspection, Textile Control clients significantly reduce the risks of receiving poor-quality goods, financial losses from unscrupulous suppliers, delivery delays and other costs for various parameters.

   Assistance in the formation of QC departments

   For textile enterprises from Uzbekistan, Textile Control provides INEXPENSIVE consulting services: assistance in turnkey formation, from consultations to recommendations on personnel selection.

   As a result of cooperation with specialists Textile Control textile production receives a functioning QUALITY CONTROL department outsourced (if it is not present in production), or - effective assistance in the formation of the QC department and training of employees for independent work in future.

   The absence of a quality control department in the structure of textile production leads to decreased sales and financial losses. The reasons for this are the lack of knowledge about the basics of QC, basic defects, GOST norms and accepted global standards, without which it is impossible to create competitive products.

   Textile Control will help you create a full-fledged quality control department at your enterprise. The formation process includes:

   • training for QC department managers and specialists
   • analysis and transfer of a database of basic defects in accordance with global standards
   • full range of consulting services.

   Possession of this range of knowledge and skills will logically lead textile production to the actual equalization of the quality of production lines in production.

   – quality guarantee in accordance with international standards!

   The working process:

   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   
   

   
   
   

   We recommend reading

   • Crochet flowers patterns with description - how to crochet a flower Crochet for beginners flowers with description
   • Confirm preferential mason experience
   • How to comb very tangled hair?
   • Women's sneakers: how to choose and what to wear with them