Qa qc testing concepts differences. Terminology: What is quality assurance and how is it different from testing?

V. P. Vorobiev , RRC "Kurchatov Institute"

Stages of development methods of quality management, throughout the history of production technology have formed a ladder of ascent to greater perfection of the dialectical link “quality management - technological development”. Perhaps it is unknown where the beginning is and what follows next. Quality management is both development and the result of development.

QC (Quality Control)- individual quality control;

SQC (Statistical Quality Control)- quality management using statistical methods;

TQC (Total Quality Control)- total quality management (technological);

QA (Quality Assurance)- quality assurance;

TQM (Total Quality Management)- total quality management (quality management);

QofM (Quality of Management)- quality of management.

Individual process control (QC). The first step in the development ladder of quality management in the production of goods and services was individual quality control among artisans, and then at the first manufacturing enterprises. Having begun in time immemorial, it was preserved even at fairly large machine-building enterprises. For example, the sliding bearings of the wheel pairs of the first steam locomotives and carriages produced at the Putilov plant in St. Petersburg were individually scraped for each pair, and it was impossible to swap wheels without additional adjustment. This was the case until serial and mass production appeared, and methods of individual control ceased to satisfy them. The search for new forms began.

“Everything new is well-forgotten Russian.”

Proverb

Corypheas. Tsar Peter I. In 2002, at a seminar on quality management in Dresden, German professor Herr. Pilz, apparently to please the Russian participants, spoke about the decrees of Peter I on control quality at the Tula arms factories. For some of the seminar participants, who had firmly learned in the last decade that there has never been and cannot be anything good in Russia, and not only in the field of quality management, this was a revelation. But, thank God, not for everyone. Your interlocutor also knew some of this history of quality management in Russia.

“All ranks in the service, as well as the owners of manufactories and other important craft establishments, should remember: all projects must be in good working order, so as not to waste the treasury and not cause damage to the fatherland. Whoever blurts out plans anyhow, I will deprive him of his rank and order him to be beaten with a whip.”

PETER

The first ships built by Peter I at the Voronezh shipyard under the leadership of Dutch “specialists” who, like many of their current colleagues from the West, came to make money on reforms and had previously only seen the ships in pictures, all of them, "overkill was committed" that is, they turned over. Russian mystics explained this by the fact that instead of the traditional Muscovite red flag, the tricolor was raised. Four years earlier, Peter I adopted it as a naval standard, converting it from the Dutch state flag by rearranging the colors. But these mystics decided that since the Russian color on the flags went to the bottom, then the ships should have capsized. The Russian shipbuilders believed that this happened because the Dutch “specialists” simply had no idea about the ship’s metacentric height (like some current economists, whether invited or homegrown, about economics, if we are not talking about the amount of their own remuneration) . Peter I made quick conclusions in both directions. He replaced the inverted flag, which, as Russian mystics said, also brought defeat to the Russian troops near Azov and Narva, with the Andreev flag, which, no matter how you invert it, “does not change its nature”, and he went to Holland for shipbuilding experience. The decree quoted above did not apply only to the fleet. For defects in the manufacture of arquebuses and fuzes at the Tula factories, he established a whole system of punishments. According to it, junior clerks assigned to monitor the quality of these weapons were deprived of a daily glass of vodka for a year, but for higher ranks of the supervising hierarchy the situation ended worse: whipping and exile to hard labor. Apparently, Peter I knew well the psychology of the then bureaucracy, which, it turns out, differed little from the current one. Since then, her psychology seems to have remained almost unchanged, except that her appetites have increased, but for some reason Peter I is not on them. And we ourselves have abandoned strict control of the bureaucracy from below. Peter I also introduced at defense factories military acceptance, sending there army engineers, whose duties included life tests, since two arquebuses and fuses from every hundred were sent to the shooting range and fired from them until they "spoiling".

Quality management (technological) by statistical methods (SQC). As long as the enterprise (economy, society) operates, quality management must be updated and continuously develop [I]. The SQC methodology was the first step in applying abstract scientific methods to it. Sometimes this technique is associated with the introduction in the twenties of the last century of statistical control in US engineering industries using the so-called Shewhart cards. The prerequisites for this were created during the search for an organization of technological processes adequate for mass production by Taylor and Ford. Ford even tried to organize quality circles at its factories. Did not work out. E. Deming managed to transfer this experience to Japan, where such circles, unexpectedly for the whole world, received intensive development.

In the 1930s, the first statistical control standards products that quickly spread to other countries. The Second World War, which required mass production of military equipment, gave a powerful impetus to statistical quality control. In the Soviet Union, in 1942, a special government commission was created, headed by one of the top officials of the state, V. M. Molotov, for the technological standardization of military production. It so happened that the tank turrets manufactured at the factories transferred to the East were not mated to the chassis of the tanks. Your interlocutor once visited the city of Gorky (now, as before the revolution, Nizhny Novgorod) in a very highly qualified experimental design bureau, which grew on the basis of such a plant. One regular worker who worked there in the winter of 1941-1942 said that some workshops at that time worked in forty-degree frosts in the open air. Construction work on them was carried out by captured Germans, who said, looking at this, that Hitler would never defeat Soviet Russia. V. M. Molotov’s commission quickly developed the MNSCHH standard (Interdepartmental Standard of the Drawing Management System). It operated until the end of the 60s, when it was replaced by its more advanced modification - the ESKD standard (Unified System of Design Documentation). This standard is still in effect, being an unsurpassed systemic regulatory document for the West.

Another wartime example is the history of the German XXII series submarine project. Separate sections of its design were manufactured at different factories, and then these “circles”, along with all the equipment located there, were welded into a whole “sausage” at the assembly plant. This technology allowed, until February 1945, through the combined efforts of factories in the cities of Kiel, Rostock-Warnemünde, Hamburg, Bremen, Lübeck, Wismar, Stettin (Szczecin), Danzig (Gdansk), Königsberg (Kaliningrad) to assemble 145 units of boats. However, they were unable to take part in the fighting. A Soviet submarine under the command of Alexander Ivanovich Marinesko, which the Germans already considered part of the Reich, destroyed the German transport Gustav Gustlov along with submarine officers who had fled from Riga and were training there. Hitler called A. I. Marinesko “enemy of the Reich No. 2.” Let us remember that “Enemy of the Reich No. I” was Yuri Levitan. And to this day, modern submarine designs are based on the ideas of the German XXII series.

Corypheas. Walter A. Shewhart proposed, during the mass production of certain parts (operations), to mark their technological tolerance zones on special maps. During the manufacture of parts, the worker noted the values ​​of the obtained dimensions on maps. These marks formed a curve showing the dynamics of deviations. Once a noticeable trend emerged, it was possible to predict when the resulting deviations might cross the warning line and when the tolerance line. Thus, it was possible to readjust the machine in advance or sharpen the cutting tool without waiting for mass defects to appear. Similar maps are still used today, sometimes they are called “traffic light”, since the zone of known good parts is indicated in green, the warning, but still acceptable, in yellow, and the defective zone in red.

Total quality management (technological) (TQC). The Second World War intensively promoted methods of statistical quality control in the United States. A trio of young energetic Americans - Edward Deming, Joseph Juran and Armand Feigenbaum - played a major role in their development. Immediately after the war, they put forward the idea of ​​extending the principles of quality management to the management of specific enterprises and the economy as a whole. But American capitalism, following the principle of “mass production of cheap goods,” saw only costs in quality management. Then these scientists applied their ideas to Japan, whose goods at that time were a symbol of junk quality. The unlimited dictator there in those years was the American General Douglas MacArthur. He decided to show these “japs” what an American engineer was, and in 1947 he invited Dr. Deming to Japan (Now this year is considered the beginning of that same Japanese way (“The Japanese Way”). However, very unexpected turns happen in history "The Japanese turned out to be not only capable students, but also smart. The Japanese Union of Scientists and Engineers (JUSE) and the Doikai Association of Japanese Industrialists believed in the promise of the proposed strategy and began to provide Deming and his comrades with all possible support in their endeavors. "Personnel decide everything!" - exactly according to Stalin, they decided and gathered a group of capable students to teach the methods of quality management proposed by the Americans. Deming and his colleagues held their first lectures, as they say, right in the hotel where he lived, and at first the students were fed. The names of the students of this groups, Matsushita and Honda, subsequently became well known in the business world. Much less know that they were still well known in the narrower circles of quality management specialists and were involved in fundamental parts of his theory. Firstly, according to the section of responsibility to society of owners and managers of enterprises and, secondly, to the section of responsibility to society of the labor collectives of enterprises.

Having learned the lesson that war is too expensive an economic undertaking, the Japanese did not abandon the idea of ​​taking a leading place in the world, but by economic means. To do this, based on quality management methods, in defiance of the “Americans,” they put forward the strategy of “mass production of cheap goods highest quality". This was a discovery at the level of creative Marxism, although its authors hardly considered themselves Marxists:

economic development through optimization of the production-consumption link through continuous improvement of the quality of goods and services. They set the task of becoming one of the leading countries not only in terms of production level, but also in terms of its quality within one five-year period. And they completed this “impossible” task, following the example of the Soviet Union, in four years. Strategy TQC was adopted in Japan as one of the main parts of the national development strategy. YASUiI took over the coordination of quality management, including holding conferences to award the Deming Prize, the first award of which took place in 1951. Ac 1960 in Japan began to raise the red flag of the Quality Month every year in November (since 1978, the same thing began to be done China every September). From then until now, Japan has held various conferences and produced more quality publications than any other country in the world. At the same time, consumers act as equal partners in them. The United States did not pay enough attention to this revolution in quality management until products labeled "Made in Japan" began to compete in the American market, a very difficult task considering the position that the American dollar received after After the United States unilaterally changed the content of the Bretton Woods agreements, in response, the Americans began to apply trade sanctions against Japan, but soon realized that this could make them even worse.

Moving from quality control to strategy TQC~ this is a transition from the process of transferring products from the category of ready to the category of good or bad, to the process of liquidation reasons marriage. Key provisions of the strategy TQC, which has largely retained its relevance for Russian enterprises even now, are:

ensuring quality at every stage of the production process in order to achieve defect-free production. However, this is not just control, it is “identification and elimination of the causes of defects;

complexity of quality management, allowing to reveal causal relationships and detect defects before they cause catastrophic consequences;

comprehensive quality management, allowing engineering services to quickly respond to changing customer requirements and wishes;

complexity of quality management, aimed at changing the consciousness of all personnel of the enterprise and allowing to filter out useless and false information regarding the quality of products and organizational processes.

Strategy TQC considers the customer as an integral part of the production process, thereby helping to overcome departmental barriers. Another main aspect is that only well-trained specialists should be present at workplaces [I]. In Japan, training in quality management methods begins in school. In particular, graduates of Japanese schools have better knowledge of mathematical statistics methods than graduates of American colleges and universities.

Having mastered TQC strategy, Japan has surprised the world for many years as a recognized leader in quality management. The goal of the Japanese five-year plan 1988-1992. there was maximum satisfaction of consumer needs, and by 2000 the Japanese were going to end the exploitation of man by man. Did not work out. And for several years now, the Japanese have been scratching their heads as to why.

TQC. Soviet experience. Engineers and scientists of the Soviet Union dealt a lot with the problem of quality, to which the very nature of socialist thinking prompted them. In the 60-70s, more and more new, as they said then, initiatives arose in this area, the most famous among which was the KANAR-SPI system (Quality, reliability, service life from the first execution). This initiative, among other things, received serious support from the military departments. But the Soviet bureaucracy, for which it was always the most important thing to report to higher structures using the balancing act of numbers rather than real results, did everything to make quality management necessary deploy like something foreign. We continue to implement quality management methods today, instead of mastering and applying them. They don’t interfere in any way, but we infiltrate them and infiltrate them, issuing menacing orders. You just need to take it and train the staff.

They say that the word “implementation” came to us from English-language regulatory documents as a translation of the term implementation. At least in the document GOST R ISO 9001-2001 and in the text ISO 9001:2000 (E) these words appear in the same places. Well, let's turn to the dictionaries. Here, for example, is the English-Russian dictionary from the publishing house “Soviet Encyclopedia”, Moscow, 1967, containing 70,000 words and expressions, compiled by prof. V. K. Muller. This is the most popular and widespread dictionary in Russia, on the basis of which other, less detailed dictionaries were compiled (and are being compiled!). When referring to it, Russian engineers and scientists usually say “Muller’s Dictionary,” as if emphasizing the professionalism of the translation. Word implementation it is translated as “implementation, fulfillment.” According to your interlocutor, such a translation is more accurate than this very “introduction”. How is this translation viewed from the “other” side? Let's look at the English-Russian educational dictionary “For persons who speak English” (“About 75,000 words and illustrative phrases”) published by Pergamon press (Oxford, New York, Toronto, Sydney, Paris, Frankfurt) and “Russian language”, Moscow. It contains a verb to implement translated as “keep a promise, fulfill an agreement”, also not bad. Where did this damn “implementation” come from? Maybe from the same Mueller dictionary? In it next to the word implementation worth the word implantation It certainly means “implementation”, even with a medical-operational bias. So what is this - a mistake by a translator who picked up an English-Russian dictionary for the first time? Or a cunning specialist who once made this substitution for purposes understandable to him? “Maybe they won’t notice!” - he thinks, hoping for a possible justification: “Just think, I made a mistake with the line!” But what benefit have thousands of quality management specialists seen in such a translation since then? Is it more pleasant for them to “implement”, while remaining innocent of anything if they “don’t implement”? Or maybe it’s time to do the difficult, but necessary and very useful “implementation” and “execution”?

Soviet achievements in the field of quality management still successfully serve Japan. She adopted our experience such as slogan calls, honor boards, communist labor teams (in the form of self-managed teams at production sites and circles for mastering quality management methods), production competition and much, much more, not to mention technological advances. However, at the same time, they completely eliminated from our Soviet experience the bureaucratic window dressing that had distorted it beyond recognition. And in Russia the names of the brilliant people who proposed these methods are unknown and forgotten.

Corypheas. Edward William Deming. His theoretical and practical work played such a role that in 1998 at the meeting of the European Foundation for Quality Management, five years after his death, 60% of the 2000 participants named him the most authoritative person in the field of quality management [3]. Over the 93 years of his life, he collected a huge collection of high awards and titles from different countries, including 18 honorary degrees and titles posthumously. However, this American scientist achieved his greatest success in Japan, where, with his help, the strategy TQC-TQM has become an integral feature Japanese way of life. And Japan deservedly noted his achievements. The Union of Japanese Scientists and Engineers established an annual Deming Prize for improving the quality and reliability of products, and the Emperor of Japan awarded him the Second Order of the Sacred Treasure Medal, which roughly corresponds to the title of National Hero of the country. Soon Japanese goods, which had previously been a symbol of cheap but junk goods, became a symbol of cheap, high-quality goods.

E. Deming's strategy - focus on performers:

1. Customer satisfaction.

3. Most defects in production occur through the fault of managers.

4. The most important thing is Human-factor (human resource management).

5. The need for continuous training.

6. Continuous improvement is the responsibility of all staff.

7. The first priority for improvement is processes.

8. Rely on facts and statistics as sources of improvement.

E. Deming concentrated his experience in "14 principles Deming". The date of their final formulation is said to be 1980 [3], but he began working on them back in the 50s in lectures for senior executives of Japanese industry. They still hold true now, more than 20 years later. E. Deming proposed them to large and small US enterprises, as well as for the transformation of the American economy as a whole, but they are still constantly discussed by quality management specialists in different countries, and in recent years in Russia. Assuming that this book will be read mainly by beginners, your interlocutor took the liberty of listing them.

1. The constant goal is continuous improvement of products, their competitiveness, and jobs.

2. New philosophy: responsibility of business leaders, their leadership in implementing changes.

3. Priority of management over control. Quality as a characteristic of goods.

4. Stopping the practice of purchasing components at low prices. Selecting regular suppliers and long-term work with them.

5. Continuously improving production and reducing costs through improved quality and productivity.

6. Continuous on-the-job training of personnel.

7. Update management methods. Checking its work, as well as the work of production departments, for the purpose of improvements.

8. Eliminating the fear of punishment for marriage and mistakes.

9. Breaking down barriers between departments. Product production from research and development to manufacturing and sales is a single process.

10. Refusal of general slogans, sermons and calls, not supported by action, to increase the quality and productivity of labor in production, since the reasons for low quality and low productivity lie in the current system, that is, outside the power of workers.

11. Refusal of quantitative indicators of production and quality planning. Changing leadership style.

12. Removing barriers that prevent employees from taking pride in their skills and their work. Refusal of annual certifications and objectivist, including digital, assessments of labor results.

13. Encouragement of education and self-improvement.

14. Improvement programs are everyone's business.

At the same time, Dr. Deming resolutely insisted that it was impossible to improve the work of an enterprise by mastering some part of his recommendations, that they all must be mastered. True, he added that an employee who begins to introduce them partially will soon be convinced of the need to use the rest.

For those managers for whom recommendations for improvement are not enough, Demiig compiled a list of them "deadly errors and diseases."

1. The fastest way to demoralize an enterprise is the duplicity of management: “Do what I say, not what I do.”

2. Lack of persistence in improving quality, desire for immediate benefits.

3. Slogans, appeals and superficial changes, instead of consistent quality management throughout the enterprise.

4. Belief that automation is the most important thing to improve quality.

5. Belief that solving current problems and buying new machines and devices will change things.

6. Re-evaluation of computerization.

7. Orientation of quality management programs towards ordinary performers and existing technologies.

8. The belief that meeting technical requirements is all that needs to be done.

9. Frequent movement of managers from place to place.

10. Ranking of personnel.

11. Commitment to the “maintaining quality level” strategy.

12. Copying other people's positive experiences instead of learning from them.

13. Treating education and training as costs rather than investments.

14. Belief that quality management can be implemented.

15. Hope for the existence of magical formulas for success, as simple as three cents.

16. Belief that the quality management department is responsible for solving all quality problems.

17. Use only quantitative criteria in production.

18. Misconceptions about the “zero defects” theory.

19. Having too many accountants and not enough engineers and statisticians.

20. The belief that “quality management consultants must understand everything about our business.”

21. Class differences and antagonisms.

22. Excessive regard for traditions or their complete denial.

23. The use of only visible factors in management with weak (or no) attempts to identify unknown or unknown ones.

Corypheas. Joseph M. Juran. Another famous American specialist who worked in the same period of time as E. Deming. The main provisions of his strategy are leadership orientation:

1. Customer satisfaction.

2. Responsibility of managers.

3. Enterprise strategy is the basis of activity.

4. The cause of most defects in production is management errors.

5. Quality management is quality planning, quality management coordination, continuous improvement.

6. Creation of specialized groups in key areas of quality management.

7. Human factor (human resource management) - continuous training.

8. Focus on facts. Statistics provide facts.

Corypheas. Arman Feigenbaum- also a famous American who worked during Deming's time. The main provisions of his strategy are orientation towards professionals:

1. Customer satisfaction.

2. Responsibility of managers.

3. TQC (Total quality control) - concept and system.

4. Quality management covers all aspects of the business.

5. Personnel training and development is part of the quality system.

6. TQC - quality planning, quality management coordination, quality improvement.

7. The basis is facts, the source of facts is statistics.

It is believed that it was A. Feigenbaum who proposed the idea of ​​integrated quality management in the early 50s, which he described in the work of the same name in 1957.

Corypheas. Kaoru Ishikawa. Many consider Prof. K. Ishikawa.

The main provisions of K. Ishikawa’s strategy (strategy TQC):

1. Constant, over many years, solution to quality problems based on modern technological, organizational and social achievements in the world. Persistent study of consumer (customer) requests, including assessment and measurement of quality indicators.

2. Quality management is a complex problem. Universal participation in quality management of enterprise personnel from top managers to cleaners. Special responsibility of managers.

3. Systematic and universal training: “Quality management begins with training and ends with training” [I].

4. Mobilization of the creative potential of staff. Quality mugs.

5. Constant updating of existing quality management methods and systems.

6. Particular attention to the organization of quality management in the workplace, where quality is created.

7. Quality management as a primary national program. State support for quality programs.

“Through total quality management, involving all workers and employees, including the company president, any company can create products...of higher quality at lower costs.”

K. Ishikawa

Like E. Deming, K. Ishikawa considered modern quality management a decisive restructuring of the thinking of enterprise managers and the work of all its employees, and not just specialized services. This attitude towards quality management is associated with a change in the formula of market relations. Was: “Market for goods.” It became: “Goods for the market.” That is, assessments of the quality of goods began to include such parameters as the predicted wishes of consumers and, even, their psychology [I]. Subsequently, this idea was reflected in the ISO 9001:2000 standard as a defining factor of modern entrepreneurship: “The focus of modern quality management is consumer (customer)".

K. Ishikawa constantly appeals to the man who, perhaps largely thanks to him, has been in first place in Japan for many years: “Quality is not given by machines, but by people.” It is known that quality often declines in Japanese factories in other countries. Then Japanese managers apply Japanese methods of working with personnel, associated with respect for the creativity of workers, which leads to an increase in their dedication to their work. This is a reflection of K. Ishikawa’s critical attitude to Taylor’s methodology, which views workers as an appendage to the machine, blindly obeying the orders of their superiors.

K. Ishikawa came up with the idea of ​​cause-and-effect fishbone diagrams (“fish diagrams”), so named for their external resemblance to a fish skeleton. Prof.'s experience Ishikawa summarized in the book “Japanese Quality Management Methods” (What is Total Quality Control? THE JAPANESE WAY) [l]. Now it looks somewhat outdated, corresponding to the TQC stage in Fig. 1. But this does not detract from the value of the book. She's helpful. firstly, to understand the philosophical essence of quality management, which, first of all, the directors must master. Secondly, although seemingly simple, it contains a lot of practical advice that remains very relevant today. One of the executives of IBM in Japan speaks about their value in the preface of the book: “Every time I re-read this book, I receive new information and new knowledge about how to manage quality.” And this conclusion is still valid.

Training in quality management methods of all categories of workers - the key thesis of K. Ishikawa’s strategy. This is what distinguishes Japanese enterprises from American ones, where quality management is entrusted to specialists and consultants. Europe and many other countries occupy some intermediate positions.

Japanese quality circles consider their main task to be teaching new techniques for achieving quality when developing new types of products. Only well-trained workers are present at all workplaces. At some companies that follow these ideas of Ishikawa (for example, Samsung), every employee from the director to the cleaner must undergo annual training, and during this time he must live in the hotel of the training center so that nothing distracts him from this part of the work process.

K. Ishikawa included personnel training in the Deming cycle. At the same time, the sector Plan he broke it into two parts, and into a sector Do I also included a training section related to new goals and ways to achieve them at this stage.

Plan: 1) determining the goals and objectives of the enterprise, 2) determining ways to achieve it.

Do: 3) education and training, 4) implementation.

Check: 5) checking the results.

Action: 6) development and implementation of corrective actions.

The cost of staff training is an investment, not a loss. Along with professional training, training in quality management methods is necessary, which must begin with the development of national and international norms and standards for quality management and criteria for the effectiveness of their implementation.

Quality assurance (QA). Statistical methods gave rise to strategy total process quality management (TQC), associated mainly with incoming, operational and output quality control. This was a big step forward, which, among other things, required the development of methods that would create confidence that the required quality would be achieved. And this has already required those involved in quality management to pay attention not only to technological functions, but also to management functions, which combine the work of all divisions of the enterprise, as well as related third-party production structures, in order to guarantee ( less responsible - to ensure) obtaining the required or declared quality. It cannot be said that this was a completely new quality management methodology. Rather, this term indicated the need for a new philosophy, according to which the final technological quality depends on all employees of the enterprise, without any exception. In addition, all efforts to ensure quality must be associated with measures to save all resources involved in production. This was soon joined by environmental problems and concern for the company’s personnel. And the main task of caring for the quality of products and services was determined - consumer satisfaction. Thus, based on the philosophy of quality assurance, the methodological strategy of TQM arose.

Total quality management (TQM). Japan's commitment to quality management in all spheres of life has absorbed Soviet experience, Ford quality circles, and much more, and began to be perceived by admiring foreign visitors as components of the Japanese way of life. The world looked at these Japanese “tricks” with suspicion for some time, but then followed suit. Thus, the universality of quality management acquired another vector - the universality of the dissemination of this strategy.

Perhaps the earliest to get involved in the development of new methods of quality management was Great Britain, which had already done a lot in this area. Since the beginning of the 90s, the rest of Europe began to intensively master TQC-TQM methods, which was greatly facilitated by the European unification processes.

Instructional and procedural technologies

As production became more complex, the emphasis in quality management changed from technological to managerial. But the technological part (metrology, regulatory and technical control, standardization) did not disappear, but it also changed, became more complicated, and improved. This division began, perhaps, with Shewhart's maps, but became quite conscious in Japan after World War II. The difference between these two areas lies in their purpose and organization. Quality management is characterized by procedural technology, while technological quality control is characterized by instructional technology. Due to the confusion of these concepts, numerous terminological and all sorts of other errors occur.

In Japan, there is an opinion that effective quality management is possible only in countries with hieroglyphic writing [I], since, by its very nature, it operates with concepts, while alphabetic and syllabic writing tends to be unambiguous, not allowing variation in execution.

In Russia, the concepts of managerial quality management and technological quality control are still very often confused. Therefore, sometimes the director (directorate), succumbing to the fashion for quality management, makes the decision: “Transform the quality control department (options: regulatory control department, standardization department, metrology department or something similar) into a quality management service.” They say that this department has existed for many years, performs similar functions, and the head there is a completely decent and manageable person who will not give away any incomprehensible numbers. The new service is often called the “Department of Standardization and Quality Management.” By the way, this was the name of the corresponding department of one important ministry. First of all, the head of this newly formed department, who was also given the title of Deputy Director for Quality, ordered that in the future all technological instructions and STP of the enterprise be called procedures. Sometime in the late 90s, one of our co-executors sent us a request: “Please send me a list of production instructions and STPs, which you now call procedures.” I can imagine how pleased the author of the letter was, who put all his abilities for humor into it.

Once, your interlocutor had the opportunity to give lectures on quality management methods to the senior and middle officers of his staff at a plant. The director, however, never came to see them, and the chief engineer only occasionally dropped in briefly. They, as Kaoru Ishikawa [I] used to say, had, of course, “more important things to do.” There was no quality service there yet, but in the most conspicuous place sat an imposing man, who all the time looked at me with a condescending smile. During the final poster discussion, he said something like this: “Everything you said here is, of course, very interesting, but it has nothing to do with us. Our quality is excellent even without all these tricks (note that this was the absolute truth). Let THEM come and have a look.” As it turned out, it was the head of the quality control department. I was never able to explain to him that in a market economy, the enterprise itself must take care of selling its products. And THEY most likely will not come to their plant themselves. And if they come, then perhaps only for the “healthy life!” obtain various know-how of this enterprise. By the way, it is precisely from this, among other things, that the enterprise’s documented quality system, implemented in accordance with modern requirements, is designed to protect. It’s good that the other heads of departments and workshops of this plant took part in the discussion of this problem animatedly and with great interest and initiative.

The commanding fathers of quality management, E. Deming, A. Feigenbaum, J. Juran, K. Ishikawa, warned us not to confuse quality management with technical control, standardization and the like. Why? Your author has a mystical respect for the workers of these services. Well, how many people are there around us who consider them second- and third-class specialists? They say, if a designer can’t resist even being a technologist, then he has only one path left - to the quality control department, standardizers or norm inspectors: “You, Kashtanka, are an insect creature and nothing more. You are against a man like a carpenter against a joiner.” In such conditions, these services gather people who are fanatically devoted to their, as they are convinced, extremely important mission, which in fact is so. But this work instills in them a special way of thinking, which very few can get rid of. Resistance to any kind of pressure on them and even sometimes insults sometimes develops due to the flexibility of thinking that quality management specialists should have.

The quality control department forces, the quality manager convinces.

The quality control department puts a barrier to defects, the quality manager looks for its reasons and ways to prevent it.

The quality control department prohibits, the quality manager encourages people to be creative.

The quality control department resists change in order to prevent defectors from opening a loophole; the quality manager stimulates continuous changes in the name of constant improvements.

But God forbid that the quality service is headed by iron-willed quality control specialists, and your author encounters this. In the position of a quality manager, such a leader begins to crush everything under himself, dictate to everyone the rules he creates, gradually driving everyone into a state of strict “order”, in which everything stops.

What will happen if the quality service is headed by a person who is able to combine the ability to provide, quality control, and the talent of a modern quality manager in one person? Then, we can say with complete certainty that he will not last long in this position. Such an outstanding person will definitely be promoted to chief engineer, or even higher. And such cases are known.

So we will adhere to the formula: “To Caesar, what is Caesar’s, and to the mechanic, what is mechanic’s.” And this dual scheme will provide us with a dialectical path to progress, depicted in the diagram that your interlocutor called the Morlin diagram.

Morlin's diagram (the art of the navigator). The transformation of the company's quality system in order to comply with modern views and international regulatory requirements should have the goal of creating a future-oriented, equally durable design. Being turned to production, in the current management hierarchy it should help eliminate outdated elements, manifesting and supporting new ones and completing missing elements.

Before embarking on such a transformation, we need to imagine the perspective from the moment of deciding to commit to a policy of continuous improvement and documenting the quality system to the time when we can say: “We have completed the big Deming cycle, we now have a complete and certified description of the quality system, its advantages and disadvantages are clear to us, and we are ready to move on to its cyclical improvement, bearing in mind that this process is endless.” To do this, we must clearly understand where we are now, and in the future we must constantly seek out what new technologies we have to master, both in the field of production and in the field of management.

This picture is represented figuratively by the Morlin diagram. It is unlikely that Kjell Morlin himself, one of the leading specialists in quality management in Sweden, from whom your interlocutor borrowed it, is aware of this name, but I hope he will not be offended by your author for such liberty.

The diagram shows the danger of extreme organizational measures. At the beginning of bringing the quality system into line with modern requirements, one should not engage in self-deception. In general, this should not be done at any stage, but especially at the beginning, when many important problems need to be solved in a short time. Otherwise, you will have a lot of free time, but only one task: how to use it.

The main thing is not to make a mistake in determining the initial coordinates, i.e. the current organizational, technological and marketing position of the enterprise. There is no need to embellish or denigrate it, we just need (!) to clearly imagine what we can firmly count on, and what needs to be given increased attention and a lot of time and money. And this “just” is a difficult matter. Secondly, you need to choose the right tack, embarking on a long journey to perfection in the stormy sea of ​​entrepreneurship, in which there are invisible underwater reefs and unexpected squalls blowing in from the wind. The most dangerous of them seem to be imperatives. For example, in the work, readers are offered a large assortment of them, namely: “must,” “establish,” “build in,” “need to try,” “carry out,” “develop and implement,” “necessary.” Well, just the deanery council. What's missing, perhaps, is "one step to the left, one step to the right." Your interlocutor, in agreement with Morlin's diagram, regards this practice with great doubt. But the author himself, however, still has plenty of fans of the Gulag style. Perhaps, now there are even more of them. But it is possible that in the past these colleagues of ours were unlucky at some point in their lives, that they were unable to develop into the role of leaders, and they are still suffering from this.

Quality of management (QofM). As we can see, the dynamics of development of quality management methods over the past 100 years is very high. And it is not surprising that a problem has recently arisen quality management(see Fig. 1). In fact, we are talking about the quality of management of enterprises and organizations of any profile and purpose in general. So, having left the management of the enterprise, quality management returned to it. In a significantly different form, though.

Most active in solving the problem From Quality Management to Quality of Management (“From quality management to quality management”) Europe addressed. Essentially, the 2000 version of the ISO 9000 series of standards is dedicated to the transition TQM to QofM. This strategy covers all aspects of enterprise activity, from coordinating the work of participants in production and distribution of products to taking into account the opinions, wishes and expectations of customers and end consumers. What will happen next, time will tell, because Quality management, once begun, never ends.

Corypheas. Tito Conti. Transition TQM to QofM It is also significant in that in quality management the use of the concept of “strategy” has moved from the realm of terminological categories to the realm of practical actions related to the formation and functioning of the TQM model. In Europe, this has been happening, perhaps, for the last 15-20 years and is largely associated with the name of Tito Conti. He is one of the main ideologists of the European version of the TQM model and the European Quality Award. This model is also called the “Excellence Business through Self-Assessment” model or the “EFQM model” (European Foundation for Quality Management). Its methodology is based on expert assessment of enterprises’ performance according to nine criteria. The predecessors of the European model were the model of the E. Deming Prize in Japan, operating since 1951, and the much later model of the Malcolm Baldrige National Prize for Quality in the USA. In M. Baldrige's model, when assessing the level of quality of companies, along with indicators of product quality, production success and customer satisfaction (as in the Deming Prize), the participation of business partners and enterprise personnel began to be taken into account for the first time. Several national prizes in Europe were established in her image. Then a pan-European model appeared, which also included an assessment of the enterprise’s responsibility to society. By focusing on such outcome indicators as customer satisfaction, enterprise staff satisfaction, community satisfaction and business results, this model is perhaps the most advanced yet. It has served as a model for a number of national quality awards, including Russia.

Somewhat later, these criteria, in the form of priority orientations, were included in the 2000 version of the ISO 9000 series standards. The strategy for the long-term success of enterprises was not the desire to grab something for certain individuals, but the philosophically long-known, but constantly forgotten truth that the struggle for any kind of primacy (material, including) or power ultimately leads to losses on all sides and degradation of society as a whole.

However, Tito Conti notes that the listed strategic goals of enterprises and the functioning of their quality systems are still largely in the realm of good intentions.

If a scientist discovers a fact suitable for publication, then it becomes the central element of his theory.

Mann's Law from Murphy's Code of Laws

Civilization factor. The authors of various publications on quality management pay little attention to this factor. But in vain. Ford quality mugs only found use in Japan more than half a century later. You can, of course, say that Ford was ahead of his time, this happened to him. But even after another half a century, outside of Japan, they somehow don’t take root very well. Or they take on another form. In the USA they are still rejected, preferring to buy ready-made technologies and specialists trained in other countries. Since, thanks to the achievements of American printing, they pay four cents where all other countries have to shell out $100, it costs them less. In other words, for some countries quality circles are more acceptable in terms of their national or civilizational way of thinking, while for others they are less so. The same can be said about many other aspects of quality management.

The undoubted merits in creating the modern foundations of quality management are universally acknowledged to belong to the American trio Deming-Juran-Feigenbaum. However, in the USA, despite their many years of efforts, the presence of the M. Baldrige Award and the unconditional successes in this field of activity in other countries, quality management occupies much less place in business life than in Japan and Europe.

Today's Russia is still diligently copying all the worst of “their” economy and no less diligently trying to preserve all the most inert of the directive methods of the economy of the Soviet period. In Russia, quality management is still associated with the production of documents that look like real ones, which “should not interfere with work.” Although your interlocutor’s foreign quality managers have repeatedly said that this period is inevitable and that it will also inevitably pass, it has dragged on too long. Nevertheless, specialists such as Tito Conti and some others have in recent years paid a lot of attention to promoting the most modern methods of quality management in Russia. Perhaps they see the greatest prospects for the development of quality management in the direction from TQM to QofM specifically in Russia because of our national-civilizational thinking? In particular, such aspects as the orientation of the results of business life to the interests of society as a whole, commitment to working “in a team”, thorough development of regulatory and technological documentation, which, however. The Federal Law “On Technical Regulation” of December 27, 2002 will cause serious damage. Like many other aspects of “technical regulation”, which, thanks to the adoption of this law, are likely to be most effectively managed by bribery.

The Magnificent Seven

1. Seven steps to excellence: QC - SQC - TQC - QA -TQM - QofM - ...

2. Seven luminaries (noted in this book, but there are others) of quality management: Peter I - Shewhart - Deming - Juran - Feigenbaum - Ishikawa - Conti.

3. 2 x 7 = 14 Deming principles.

4. “Deadly” mistakes and illnesses of business managers according to Deming.

5. From instructions to procedures (description of processes).

6. The quality control department forces, the quality manager convinces.

7. Between the Scylla of utopia and the Charybdis of the deanery.

Literature

1. Ishikawa K. Japanese methods of quality management, - M.: Economics, 1988.

2 Adler Yu. P., Shper V, L, Origins of statistical thinking. // Quality management methods. - 2003. ~ No. 1.

3. Stein Smaaland. Doctor Wu Edward Demin - the father of the quality revolution. // European quality - 2002 - No. 3.

4. Mafsushita. Not for Breath alone. - Kyoto.: PHP Institute Inc 1984.

5. Edward Deming. Way out of the crisis. - Tver.: Alba, 1994.

6. V. Ya. Vorobiev. How not to waste money. // Quality management methods. - 2001. - No. 4.

7. V. P. Vorobiev. How not to waste money. Practice notes. - Development and certification of quality systems 2001.

8. O. V. Malyshev “Moment of Truth” in the quality management system project. - Quality management methods 2003

9. V. Yu. Ogvozdin. Quality control. Fundamentals of theory and practice. - M.: “Business and Service”, 2002.

10. V. A. Kachalov. Encyclopedia of errors in quality management. - Standards and quality. - 2003. - No. 1, 2.

11, Tito Conti. Stakeholder system: strategic value. // Methods of quality management - 2003- No. I.

12. American M. Baldrige Award // All about quality. Foreign experience, No. 20. - M.: NTK "Trek", 2000.

A professional view on the organization of quality control of testing and laboratory work from Olga Almendinger, head of the MICROMINE Consulting Services division.

QA/QC

The abbreviation QA/QC consists of two important unequal parts. As defined by the International Organization for Standardization ISO 9000 (ISO 9000 2000 Definitions):

Quality assurance (QA) is defined as a set of activities whose purpose is to demonstrate that an entity meets all quality requirements. Q.A. activities are carried out in order to inspire the confidence of both customers and managers, confidence that all quality requirements are being met.

Quality assurance (QA) – This is a set of activities whose purpose is to clearly demonstrate that the object meets quality requirements. Quality assurance activities are aimed at providing confidence that quality meets requirements for both managers and consumers.

Quality control (QC) is defined as a set of activities or techniques whose purpose is to ensure that all quality requirements are being met. In order to achieve this purpose, processes are monitored and performance problems are solved.

Quality control (QC) is a set of actions or techniques whose purpose is to obtain information that the product meets quality requirements. To achieve the goal, all processes are monitored and identified problems are resolved.

In other words, Quality Assurance ensures that a process is done correctly and produces a predictable result, while Quality Control ensures that the product meets a specified set of requirements.

A quality assurance program is typically a written policy that describes, at a minimum, the sampling, sample preparation, and analytical processes in conjunction with quality control protocols.

Thus, the quality assurance program is broader than quality control and has, to some extent, redundancy, which allows us to guarantee the quality of work (subject to confirmation by quality control data)

Quality Assurance Program

Regulatory authorities provide general guidelines and definitions for testing programs and laboratory activities that do not prescribe the use of specific methods or sequences of activities. It is an extremely thankless task to develop a list of mandatory requirements for a quality control program for each stage of exploration, type of mineral and various economic scenarios. A Competent Person or Competent Person has great freedom of professional judgment.

On the one hand, this approach is flexible and allows, if necessary, to easily adapt the work program taking into account new technologies, economic changes, etc. On the other hand, the lack of detailed information in the standards creates difficulties for mining companies in drawing up a quality control program that should be effective and meet regulatory requirements.

In fact, a quality assurance program is a work schedule + a quality control program. The quality assurance program should cover all key aspects at every stage of exploration, from drilling and sampling to the delivery of results from the laboratory. The work methodology at each stage must meet the assigned tasks, it is necessary to use appropriate equipment that will ensure the required level of quality of the result obtained, analytical methods are selected taking into account the type of mineralization, the documentation and data storage system ensures high-quality and complete collection of geological information and easy access to data. It is also important to identify specific performers whose responsibilities include monitoring compliance with the regulations for the actual performance of work and the results of the quality control program.

Such a program would include, but not be limited to, the following:

• Checking the correctness of data entry. The best control option is double data entry, when the most important information is entered by different performers and then cross-checked against two sets of data. A simpler alternative to such checking is regular checking of a representative part of the data (at least 5%) using the same method.
• For data received in digital form, it is necessary to set up a procedure for importing data directly from the device, which will avoid errors.
• Details of sample preparation and analytical methods used, including sample volume and methods for sample reduction,
• Use of duplicates/forms/standards, frequency of evaluation of results, acceptable limits and actions if problems are identified.
• Frequency of data acquisition and 3D geological interpretation.

To be sure that the developed program complies with the standards, you can seek advice from competent specialists, you can use published international reporting reports as reference information, selecting deposits with a similar type of mineralization and use the regulatory requirements of the State Reserves Committee, which contain a lot of useful information.

Methodology

When developing a quality control program, it is necessary to take into account the features of the chosen methodology. The choice of technique is determined by the size, internal structure and variability of the distribution of the useful component. Two main approaches:

• The methodology complies with industry standards.
• Non-standard methods

Choosing a standard methodology that has been proven to produce consistent results of good quality in similar types of deposits makes life easier. As a rule, such methods contain a margin of safety, which gives confidence in the result obtained. Since standard methods have been proven, they do not require excessive control and standard quality control programs can be used.

The choice of non-standard methods may be due to various reasons. This may be specific raw materials, for example, technogenic deposits or non-standard mineralization - deep-sea nodules. For these types of tested standard options may simply not exist, and choosing and debugging a technique will require more effort and time.

Another reason for choosing a non-standard methodology may be the use of historical data. From a historical perspective, the use of the technique was justified, for example, the exploration target was areas of rich vein mineralization, but using this data to estimate poor vein mineralization in the intervein space may give an incorrect result, even if historical quality control results show a good result. But using historical data and assessing the quality of such data is a separate topic.

If the choice is made in favor of non-standard methods, the quality control program must take this into account. The program must be adapted to a specific methodology and have a margin of safety that will ensure a guaranteed high-quality result.

It is unacceptable to choose a non-standard technique that could potentially give a better result, but because of its higher cost or execution time, save on the control program or other work that affects the quality of the result.

Quality Control Program

Any measurement is erroneous, the only question is how large the error is and whether it can be neglected within the framework of the task at hand. In this case, by error we mean the inconvergence and inaccuracy of data. This is not an error in the everyday sense of the word, but a result of the physical limitations of the representativeness of a small sample, designed to reflect a much larger volume, and the sensitivity of analytical methods. There can be many sources of error, such as sample heterogeneity, contamination, data uncertainty, and analytical precision.

The error associated with each source is cumulative.

Each project is unique and is characterized by its own distribution of error sources, so the quality program for each project will differ in completeness.

A full-fledged program is designed to monitor all fundamental aspects of the testing sequence in order to control and reduce the overall resulting measurement error:

It is certainly important to identify and measure all sources of error, but we are limited by financial and practical constraints. The first step in the process is to determine what needs to be measured, how often, and what to do with the results.

Operational control over these aspects is achieved by adding various control samples to the sequence of routine samples.

Each sample has its own specific purpose in the quality control protocol. The use of control samples is also useful for identifying problems with sample labeling during sample collection and processing.

The quality control program is developed individually for the project based on the quality of the working laboratory, the range of economic contents of the useful component, the distribution of mineralization and other features.

Testing

Once we have determined what we measure, we need to determine how we measure it. Each stage has its own control features. At the testing stage, important parameters are the correspondence of the sampling interval and sample volume to the type of mineralization, the quality of sampling and labeling of the sample, the quality and completeness of documentation.

Testing control

Control is carried out using control samples selected in the same way as the main ones - twin sample.

• Core halves
• Parallel groove
• Field duplicate (RC or drill and blast wells)

The result from field duplicates demonstrates sampling convergence and natural variability in mineralization. 1 duplicate per 25 or 50 privates.

To reduce errors during testing control:

• A control sample is taken as close as possible to the main sample simultaneously with the main one (by the same performer), in the same volume and in the same way and in encrypted form is sent to the same laboratory
• The sample must be processed by the same person, analyzed in the same laboratory, using the same method and in the same batch as the main sample.

In English terminology, they try to avoid the term “duplicate” for half a core or a parallel groove. Half of the core will be characterized by a higher dispersion compared to a sample taken from the cuttings of an RC well, since it has a small but physically measurable spatial displacement, which increases the dispersion due to geology, and not the sampling method.

Analysis of the second halves of the core can show whether there is a systematic error during sampling. Some geostatisticians use half-core analysis to determine the level of variance in variography as an independent estimate of the nugget effect at close ranges.

Twin wells are separated by an even greater distance and it is often difficult to evaluate the results from such wells.

Sample preparation control

Sample preparation is often the most vulnerable stage of testing, where there is a high probability of error.

Transporting samples to the laboratory is quite expensive, so sample preparation is often carried out independently in laboratories organized at the site of geological exploration.

Large accredited laboratories have their own internal quality assurance and quality control protocols; in the case of organizing a sample preparation laboratory at a drilling site, all responsibility for quality assurance falls on the subsoil user.

In Australia and Canada, large laboratories such as ALS provide on-site sample preparation services with trained personnel. Currently, such an opportunity has appeared in Russia.

Blank samples must obviously not have significant contents of the element for which the analysis is being carried out. Blank samples of stone material allow you to control the possibility of contamination of the sample with contents from previous samples during the sample preparation process. Duplicates of coarse material are collected at each sample reduction step to determine the degree of variability associated with the sample preparation and sample reduction process.

It is also useful to monitor losses and quality of crushing and abrasion and enter them into the database. If control results show an unacceptable level of convergence, this information may be helpful in determining the reasons for poor quality.

Sample control

Blank samples of stone material are inserted into the sample sequence before sample preparation and processed in the same way as the main samples.

The mineral-free nature of a blank sample of stone material must be confirmed by analytical results, but for stone blanks this is not as critical as for certified standard samples.

Blank samples of rock material must be sufficiently hard and of suitable size to ensure effective abrasion of any contaminant material that may have remained in the equipment from the previous sample.

If possible, the documentation geologist should identify more or less consistent potentially mineralized intervals and insert blank samples of rock material within or immediately after the interval. The average volume of control samples is 2%, but the amount can vary significantly depending on the project. For example, in the presence of large gold, the risk of sample preparation contamination increases, so it is recommended to increase the number of blank samples to 5%.

Duplicates of rough material (coarse-grain duplicates) are taken at the stage when the sample is reduced for the first time. This usually occurs at the stage of crushing the sample to a size of 10 mesh (2 mm). These samples are then processed in the same way as the main samples and analyzed in the main laboratory in the same laboratory batch. With their help, the quality of sample preparation and sample reduction is controlled.

After sample preparation, samples ground to 200 mesh are sent to the laboratory.

Control of analytical work

When choosing a primary laboratory, there are several factors to consider, including: quality of work, cost, convenience and range of services provided. The main laboratory must be selected based on the price-quality-deadliness ratio. The accuracy of the result may be slightly lower than that of an external laboratory, but nevertheless sufficient for use in assessing resources and reserves without the use of any corrective factors.

An adequate level of accuracy can be taken as a deviation from the true value of ±5%, which is determined from standard samples of certified material and from the results of re-analysis in an external laboratory.

For a bankable feasibility study, it is desirable that the main laboratory is not affiliated with the control laboratory and has no financial interest in the project.

When choosing a laboratory, the cost of work and service should not play a decisive role. Quality must prevail.

To assess the quality of a laboratory's work, a batch of standard samples can be sent to several candidate laboratories. It is necessary to send at least 5-6 samples in two different batches, which must be analyzed on different days. This number is not enough for a full assessment; as a rule, at least 30 results are required, however, this will give an assessment of the quality of the laboratories’ work as a first approximation. There is a special technique for ranking laboratories based on a small sample of data.

When controlling quality, laboratories use:

Control of material homogeneity

Sample heterogeneity can be a source of significant error during analytical work. It is recommended to sift about 10% of samples. In the case where the laboratory receives worn-out samples, this can become a counter-argument for the laboratory if the result is unsatisfactory. If abraded samples are transported over a significant distance, they must be subjected to repeated abrasion before sampling. It is recommended that samples pass through the grinder for 10-20 seconds before sampling. This will reduce segregation, which can occur due to the settling of denser particles, segregation of particles by shape, size, due to clumping, etc.

Samples taken for abrasion quality testing must be collected and sieved before re-abrasion.

Material used for wet screening control should never be returned to the sample. For this purpose, 10 g of material is selected from a pre-mixed sample in a checkerboard pattern.

Convergence control

Laboratory duplicates, re-analyzed in the main laboratory, allow you to control the convergence of the laboratory, laboratory duplicates sent to an external laboratory allow you to monitor the presence of drift in the work of the main laboratory.

Laboratory duplicates of abraded material, reanalyzed in the main laboratory, can be of two types:

• Repeatassay– in the same laboratory and in the same batch as the main laboratory sample. Show convergence within a batch
• Pulpre-assay– in the same laboratory, but outside the original batch. Shows the convergence of the laboratory as a whole and the degree of variability of the result over time.

It is useful to evaluate these two types of results separately because they can provide additional information about the quality of the laboratory's performance.

Standard samples

Ideally, a standard sample should have a matrix similar to the deposit rocks, have a very high degree of homogeneity, cover the grade range present at the deposit, and have a documented history of quality preparation and certification. When selecting reference materials, it is also useful to be guided by geological and metallurgical considerations.

It is useful to select a set of certified samples that would include several different levels of content in combination with different matrices.

Standard samples can be:

Commercial reference materials of certified material(CRM). Certified material must comply with ISO 9000 standards and be accompanied by a certificate. The matrix of the standard must correspond to the type of mineralization and composition of the host rocks. Ready-made standard samples are currently available for purchase. The disadvantages of such standards include high cost. The advantages of using such standards include the fact that they are widely known, which on the other hand can be a disadvantage, since commercial laboratories are also familiar with them. When using commercial standards, it is important to consider the suitability of the standard material for the mineralization composition. The most widely used standard material is produced by the New Zealand company RockLabs and the Australian company Ore Researches and Exploration. The Ore Researches and Exploration company offers a large selection of different samples and a fairly convenient service for selecting a suitable sample on the company’s website. The certificate of reference materials contains detailed information about the mineralogical composition, type of mineralization and the deposit from which the material for the preparation of standards was taken.

Standard samples of our own production (in-house Standards). If it is not possible to use ready-made commercial standards, they can be prepared independently. To make standard samples yourself, you need to determine the characteristics of a set of standards, since it is better to use one set throughout the exploration than different standards at different times. It is then necessary to determine the level of homogeneity of the standard material. This step must be completed before defining the content in the standard. This is usually done by taking 24 samples at random or by dividing the material in a representative manner and sending the selected samples to a high-level external laboratory. The purpose of the test is not to determine the content in the sample, but to confirm or not that the standard was prepared with high quality and the material is homogeneous. As a rule, if the relative standard deviation (standard deviation/mean) exceeds the value declared by the laboratory for a given method and content, then the homogeneity of the standard is insufficient and additional processing of the material is necessary. The content of the standard that has passed the homogeneity test is determined by an additional certification program.

It should be taken into account that not all deposits have material suitable for preparing standard samples with a sufficient level of reliability. Especially deposits with large gold. Since the main purpose of standard materials is to determine the degree of convergence and accuracy over time, it is not reasonable to manufacture standard samples from this kind of material. In some cases, sieving the material through a 100 mesh screen or finer and reducing the sub-screen fraction to 200 mesh may improve the situation. But in such cases, it is more reliable to use commercial standards, even if they have a different mineralogical composition, or to prepare samples from material from another deposit. In this case, it is better to control the problem of large gold by using duplicates, and the accuracy of the analytical result using a quality standard sample. Standard material with a high degree of heterogeneity is much less effective for determining drift in the laboratory.

Including Standard Samples

The low-grade mineralization standard should have a grade close to the accepted cut-off grade (for gold this is in the range of approximately 0.4 to 0.8 g/t). The rich standard must have a grade above the estimated cut-off grade of the ore being processed or in the region of the 85th percentile of all ore samples analyzed. Analytical accuracy for these grade ranges is important as it can minimize error in material classification.

Another rich standard may be close in content to the median value of all ore samples analyzed. Or to the median values ​​of oxidized and primary ores at the deposit, if different types of ores have significant differences in content.

It is also important to include a blank standard to control contamination.

Best practice is to use 4 types of standards:

1. Blank (blank standard)
2. Low grade standard - in cut-off grade area
3. Standard with average content
4. Standard corresponding to high contents of rich mineralization

To determine the number of control samples, it is important to know the number of samples in the laboratory batch and how the number of samples in the batch varies depending on the analytical method if more than one is used. This figure varies depending on the laboratory. Standards are included in the sample sequence so that each laboratory batch contains at least one ore standard, one blank, one lean standard, and one repeat. Each submission must contain several duplicates of rough material.

Inspection is often more focused on checking high-grade mineralization, but there is now a significant reduction in cut-off grades and a good result for low-grade ores is essential to reduce the risk of ore-to-rock misclassification. Therefore, it is important to pay more attention to the quality of analytics at low grades.

Blank samples of coarse material must be inserted in such a way that at least one blank passes through sample preparation per shift.

There is some agreement between consultants on the total volume of control samples. Almost everyone gives a value around 20% of the total number of samples. This 20% has now become the de facto industry standard. However, the breakdown by test sample type is not as structured. A detailed quality control program should include all types and subtypes of control samples so that precision, accuracy, and possible contamination are properly monitored and assessed at all stages of testing. The quality control program should be tailored to the specific needs of the project and the size of the laboratory batch.

It is recommended to keep the total number of control samples in the region of 20%, but distribute the number of samples by type, taking into account the problems that are most likely to occur on a given project. As the project progresses and these problems are identified and corrected, the absolute and relative quantities of control samples can be adjusted accordingly. The table provides starting numbers for control samples from which you can begin to plan your quality control program.

It should also be noted that sending samples for external control should also be accompanied by the inclusion of standards and duplicates so that the convergence, accuracy and possible contamination of the control laboratory can also be assessed independently.

The quality control program adds 15% to the cost of analytics, which is about 1-2% of the total project cost

Results Evaluation Criteria

An important aspect of the quality control program is the determination of acceptable limits for the results obtained from control samples. Assessing results across laboratory batches provides the opportunity to remove a subset of results, rather than a large set of data, if data quality issues arise. In real practice, it is often difficult to control laboratory batches outside the laboratory. The geologist responsible for the project may not know where one laboratory batch begins and ends and must disclose control samples so that the laboratory can identify the batch. Sometimes, the geologist may instead request to re-analyze a small group of samples (5-10) that included the standard sample, so as not to reveal its position. The laboratory must explain to the project geologist the problem that caused the error—mixed samples or batches—and correct the error by reanalyzing the samples.

The main idea of ​​the batch-by-batch approach is that if a laboratory batch was processed in violation of the protocol, then the entire batch is rejected. This may be true, but not always. The most common type of error is a random error associated with incorrect labeling of the sample, incorrect reading of the instrument, rearranging numbers when recording the result, spillage of material or boiling of a single sample, etc. The most common error is the mixing up of two samples, which leads to the fact that two neighboring samples have incorrect results.

It is helpful to discuss the quality of the result with the laboratory and agree on the acceptable limits of the result, how they will be determined, and what actions need to be taken to correct the problem. The details of the agreement may vary, but as a rule, laboratories will agree to retest a batch that fails inspection free of charge.

However, if the quality control program comes with problems on the customer’s side (very often there are errors in the numbering of samples and standards, poor quality of sample abrasion), then such an agreement is devalued.

Blank samples

Blank sample of abraded material (PulpBlank)-The analytical result must be less than or equal to two values ​​of the detection threshold of the method used.

Blank sample of stone material (CoarseBlank) The analytical result must be less than or equal to three values ​​of the detection limit of the method used

The absolute value of the lower detection limit must be taken into account. Some laboratories use assays with gravimetric termination, for which the certified lower detection limit is 0.2 g/t. In this case, a more reasonable option would be to use the lower detection limit of a more standard method, such as atomic absorption fire assay, which has a lower detection limit of 0.01 g/t.

Duplicates

To assess the significance of the difference in content between the duplicate and the original sample, the Relative Paired Difference (RPD) method was used.

PulpDuplicate)– the desired variation should be less than 10% RPD for 80-85% of samples for an internal duplicate within the same batch or laboratory, or 15% between different batches or different laboratories.

Duplicate rough material (CoarseDuplicate) – less than 15% for 80-85% of samples

To evaluate the results from core halves, pairwise comparison usually gives an ambiguous result. In this case, it is useful to compare the distribution of the two samples. For these purposes, it is useful to use quantile-quantile (Q-Q) plots.

When calculating RPD, results below the detection threshold are assigned a value of zero.

For pairs of samples with an average value (“0.5∗(x original + x duplicate)”) less than 15 detection threshold values, a wider discrepancy interval is allowed:

For duplicating worn-out material– the result is acceptable if:

< = два значения нижнего порога обнаружения

For a duplicate of rough material – the result is acceptable if:

< = три значения нижнего порога обнаружения.

CRM

The most common method for evaluating results is to use three standard deviation values ​​as acceptable limits.

• The standard deviation (SD) value is stated on the standard sample certificate and gives an indication of the level of accuracy expected from a controlled laboratory. The SD value takes into account the error associated with measurement inaccuracy and the heterogeneity of the material of the standard sample itself.
• The standard sample should be characterized by the amount of dispersion associated with the heterogeneity of the material, which is insignificant compared to the measurement error, which can be neglected.
• The SD value includes all measurement errors: interlaboratory variance, precision error, and standard sample variability.
• The standard deviation calculated for analytical samples produced by Ore Researches and Exploration is calculated from the same data as the certified value of the sample, obtained as a result of the interlaboratory certification program and accepted as correct.

Rejected values ​​and extreme values ​​beyond 3 standard deviations are removed from the data set. Extreme values ​​are excluded only if the homogeneity of the standard sample material has been previously confirmed and independently of this program and these values ​​can be attributed with a high degree of confidence to an analytical error and not to the heterogeneity of the standard.

Table 4 of the standard sample certificate shows options for acceptable limits.

The second method for evaluating a benchmark uses a window + 5%, calculated directly from the certified value of the standard. For reference, the table shows the values ​​of three relative standard deviations (1RSD, 2RSD and 3RSD).

For standards with contents close to the lower limit of detection, limits should be used with caution, since confidence intervals calculated from the standard deviation value may be too wide, while the limits determined by the window method + 5%, on the contrary, is too narrow.

The SD specified in the reference material certificate is calculated based on data from an interlaboratory control program in which world-class laboratories participate. In an ordinary laboratory, the error of the laboratory result may be greater than in world-class laboratories

To provide a more grounded SD value, the reference material certificate provides a summary SD value that takes into account inter-laboratory measurement errors. This “one-dimensional” approach must be taken into account when assessing the results.

Another method for evaluating results from reference materials is to use the eigenSD value obtained from a specific reference material in a controlled laboratory.

This method is offered for use by Rocklabs. Rocklabs Standard Sample Certificates do not provide an SD value that can be used to evaluate the test results, but do provide an Excel evaluation template that can be downloaded from the Rocklabs website. The principle of calculating SD is similar to the approach used by Ore Research Exploration to calculate SD under the interlaboratory control program, but using data obtained in a controlled laboratory.

Using the template is quite simple; you need to select the type of standard template on the first page and copy the data into the corresponding cells of the template. After this, extreme values ​​will be highlighted in orange in the template. The criterion used is a deviation of 40% from the sample median. Statistical parameters are calculated for the total sample and for the sample with extreme values ​​excluded. On another sheet, a control graph is constructed, the sample with extreme values ​​excluded is determined by the value of the standard deviation, which is taken as acceptable limits, samples lying outside three standard deviations are determined, and the total percentage of samples that fail the control is estimated. When new data is received, you can simply add it at the bottom of the columns and the template will automatically recalculate the result for all data.

This is what a control graph looks like, with extreme values ​​highlighted in orange and samples outside three standard deviations highlighted in yellow.

This is what the data page and evaluation result look like.

Below are the parameters by which the quality of analytics is assessed.

The main inconvenience is that everything is in English and the template is limited to 150 records; the template can be adjusted to process more. But if necessary, a similar procedure can be done in another program.

Laboratory drift

When monitoring laboratory performance, it is important to plot the average content for each of the standards used on the project over time.

Similarly, it is useful to plot the difference between the average of the primary and external laboratory results for each submission over time.

These two graphs allow you to monitor the possible presence of drift in the laboratory. If problems were identified as a result of the control and part of the data was re-analyzed, then these rejected data must be replaced with new ones before calculating the average.

A sustained difference, determined over a time period covering several shipments, of 5% or more is generally considered unacceptable.

The laboratory should be informed of the presence of drift, but it is advisable not to disclose confidential information about the reference materials used. It is better to cite data from an external laboratory as an argument, and it is recommended to disclose data on standards as a final argument.

Graphical representation of QA/QC results

It is recommended to visualize quality control data and update it whenever new data is received. A graphical representation of data, when done well, summarizes the history of the entire project and provides useful context for the current outcome. Personal preference plays a very important role in the accuracy of the presentation of control data in such graphs.

Next, we will look at options for graphical representation of quality control data. The most common type of control chart is a chart on which values ​​are plotted according to time standards. It is usually performed in the form of a line graph, so that the results are sorted by the time of analysis. On the graph, the lines show the expected content value, upper and lower limit. If during work there are long breaks in analytics, changes in work methods, etc., it is useful to display this kind of information on a chart.

If you normalize values ​​by standards by bringing the standard deviation to unity, you can display the results for all standards on the same scale on one graph. This may allow the overall drift of the laboratory to be assessed.

Additionally, you can highlight different laboratories (if they have changed) or different types of standards by color, so that you can visually assess whether the result depends on the type of standard or whether the trend is present for all types.

Scatter plot

A scatter plot is another useful way to visualize data. On such a graph, the control testing data is plotted along one axis, and the main result is plotted along the other. For information, it is necessary to plot the x=y line and acceptable limits on the graph.

The graph can be used to evaluate the results against duplicates.

After excluding extreme values ​​(visually or statistically), the most optimal regression equation and correlation coefficient are calculated.

Before calculation, it is necessary to exclude extreme values ​​(the highest or 1-2% of the population having maximum abundances) as well as values ​​close to the lower detection limit, for which the discrepancies will be the highest. This will give a more representative estimate of bias compared to the overall estimate, where a large number of zero or extreme high values ​​will have a significant impact on the regression equation. It is important to use correlation coefficient and regression line values ​​with caution because they are based on the assumption that the data is normally distributed.

It is useful to accompany the graphical presentation with basic sample information and basic statistical parameters.

RPD chart

Another useful graph characterizing convergence can be constructed from sorted relative pairwise divergence values ​​and ranked by percentile.

Duplicate worn material (PulpDuplicate)– the desired variation should be less than 10% RPD for 80-85% of samples for internal duplicate within the same batch or laboratory or 15% between different batches or different laboratories

Quantile-quantile plot

A quantile-quantile plot or QQ plot allows for a visual comparison of two distributions. It is useful for assessing the result of the second halves of the core and the parallel groove, and is useful for estimating bias.

The presence of a systematic error can be determined if the graph lies below or above the XY line. If the graph lies close to the x=y line, the distributions are similar.

Nelson's rules

Literature:

1. Lynda Bloom Analytical Solutions Ltd. Developing Effective Procedures for Mineral Assays/Analyses
2. Scott D. Long, Dr. Harry M. Parker, Dominique François-Bongarçon. ASSAY QUALITY ASSURANCE-QUALITY CONTROL PROGRAM FOR DRILLING PROJECTS AT THE PRE-FEASIBILITY TO FEASIBILITY REPORT
3. Armando Simon Mendez AMEC International Ingenieria y Construcciones Limitada Chile A Discussion on Current Quality-Control Practices in Mineral Exploration

We love adding talented new recruits to our ranks. And during the interview they noticed more than once that when asked about the difference between Quality Assurance, QA), quality control (QualityControl, QC) and testing (Testing) answers vary greatly. Sometimes they even cause heated debates with “turning the tables.”

But the point is that these concepts are defined in each company and even team in its own way. At different times, the term “testing” meant different things, which is where misunderstandings sometimes arise. We don't need such nebulae, so let's put everything in its place and figure out what's what. Hooray!

Essentially, testing and QC are part of QA, so the simplest comparison would be an ordinary nesting doll. Quality assurance is a set of activities covering all technological stages of software development, release and operation to ensure the quality of the product being released. To put it a little more simply, this is the brains of decision-making in product quality assurance teams, our biggest nesting doll.

The quality assurance process consists of:

• Checking software specifications and requirements.
• Risk assessments.
• Planning tasks to improve product quality.
• Preparation of test documentation (regulations, approach, test plan, checklist), test environments and data. Compared to QC and testing, at this stage an effective model and sequence of various product tests are developed, tools and scripts are described that will provide the necessary level of functionality coverage.
• Testing and verification of requirements and specifications.
• Product testing process.
• Analyzing testing results, drawing up reports and other acceptance documents.

The QA manager must understand exactly at what point the tester will join the project, and have time to prepare a test plan, test documentation, and environment by this time. In addition, he must have a couple of skills of other team members:

• From a marketer - understanding the target audience and market.
• From the programmer - at least a superficial understanding of the code and technical limitations for implementing the functionality.
• From the PM - a holistic perception of all parts of the project, understanding of the timing, stages and iterations of the project life cycle.

Thus, we found out that QA, in addition to direct testing and assessing the quality of a product, represents a series of organizational activities for planning and developing an approach, as well as preparatory activities. All this allows us to achieve high quality of the product, artifacts and the entire process of involving the testing team.

Quality control

Inside the QA doll is a QC. This is a check of the current state of the test object using criteria such as:

• The degree of readiness of the product for release.
• Compliance with the requirements.
• Compliance with the declared level of project quality.

Thus, the main area of ​​work of a QC manager is the quality of intermediate and final development results. This is generally controlled like this:

• The functionality is checked for compliance with the requirements.
• Documentation is analyzed for compliance with writing, content and format standards. You can check both test documentation and specifications, as well as the project schedule.
• The code is reviewed for compliance with programming standards, architectural documentation, security requirements, etc.

That is, the goal of QC activities is to provide relevant and timely quality profiles based on various methods for calculating it, depending on the stage of product development and the number and priority of defects found.

Testing is a check of the compliance of the created product with the requirements, carried out by analyzing its operation under special conditions selected in a certain way.

The general testing scheme looks something like this:

1. The tester receives the product and/or requirements at the input.

2. He creates tests and observes the behavior of the program under certain conditions.

3. The tester receives data on compliance and non-compliance of product behavior with specifications. After which he documents this in the form of a description of defects and filling out test documentation.

4. The information obtained is used to improve the product or change requirements by making modifications to the code.

During the testing process, a specialist controls the execution of the program and thereby creates conditions for observing the behavior of the product, comparing reality with the expected situation.

He knows how to determine what caused the error, or at least knows where to look for it. Standard practice includes using auxiliary tools and internal capabilities of the development platform, monitoring application logs, and working with the database.

Let's sum it up

We believe that perspective is important for quality products. Writing beautiful code and testing is wonderful, but it is an experienced QA manager who will be able to see the reasons for missed deadlines, customer dissatisfaction and, of course, a screwed-up final product or service.

And since now you know how to distinguish QA from QC and testing from both of them, you have every chance of creating level 80 software. Today and always!

In a broad sense, quality control is the sum of all measures to ensure a stable level of quality of manufactured products. In a narrow sense, this term means a comparison of the actual value of a product with a given value, in which it is established to what extent the products satisfy the requirements established for them.

Quality Control- any planned and systematic activity carried out at a production enterprise (in a production system), which is implemented to ensure that the goods produced, services produced, processes performed meet the established customer requirements (standards).

In accordance with the ISO 9000:2000 standard, which defines all such standards, quality is a set of certain characteristics and properties of a product or service to satisfy specified needs. This definition turns quality into a value-neutral list of product characteristics (see Diagram 1). It is important that the selected characteristics are measurable and controllable. These may include physical quantities (weight, temperature, density), as well as characteristics relevant to trade (price, quantity per lot, package size) or to customers (for example, positive consideration of wishes). Characteristics can be very different, two main subgroups are qualitative (for example, design) and quantitative (stroke height), each of which can be determined either precisely (for example, the piston stroke of a press is exactly 150 mm) or have a certain interval (the piston stroke of a press installed in the range from 20 to 100 mm). In addition, there may be tolerances (150 mm plus minus 0.1 mm).

Diagram 1. Example of a quality concept for a connecting hose.

Quality control includes both design (design) control and manufacturing inspection, which may differ in the volume of control activities carried out during continuous control and the sample size during selective control. Sampling control (statistical) gives indications about the state of the production process either using statistical methods (production control) or using data obtained on the proportion of defective products in the volume of the production batch.

Types of quality control

Thus, a distinction is made between sample, continuous and statistical types. Solid All products undergo inspection; production records are kept of all defects that arise during the manufacturing process of the product.

Selective— control of a part of the product, the inspection results of which apply to the entire batch. This type is precautionary, hence it is carried out throughout the entire production process in order to prevent the occurrence of defects.

Incoming control— checking the quality of raw materials and auxiliary materials entering production. Constant analysis of supplied raw materials and supplies allows us to influence the production of supplier enterprises, achieving improved quality.

Interoperational control covers the entire technological process. This type is sometimes called technological, or current. The purpose of interoperational control is to check compliance with technological regimes, rules of storage and packaging of products between operations.

Output (acceptance) control— quality control of finished products. The purpose of final inspection is to establish compliance of the quality of finished products with the requirements of standards or technical specifications, and to identify possible defects. If all conditions are met, then delivery of the product is permitted. The quality control department also checks the quality of packaging and the correct labeling of finished products.

7 instruments

The following quality control tools are available ( ):

• Summary map of defects;
• Bar chart;
• Quality regulation card;
• Brainstorm;
• Correlation diagram;
• Pareto chart.

Closely related to the technically oriented quality control is the economically oriented approach. Technical parameters should never be considered separately from economic ones. Technological innovation occurs precisely where economists see a good opportunity to reduce costs or great potential to increase profits. The potential for improvement can only be assessed when a clear economic analysis is available together with the technical data. The international standard ISO 9000:2000 defines quality costs as “the costs incurred to ensure the desired quality and to convince the consumer that the product will satisfy his needs, as well as losses from insufficient quality.” Diagram 2 gives an idea of ​​how they are divided:

Scheme 2. Structure and classification of quality costs

The cost of a defect is determined by whether it was discovered in production or a consumer complaint. Typical internal costs of marriage are:

• waste, defective products;
• recycling of defects;
• unplanned sorting;
• research of the problem;
• repeated inspections;
• additional time costs due to the need for unforeseen control.

Typical external costs of marriage are:

• costs of replacing defective goods
• maintenance and repair of defective goods
• expenses arising from the provision of a guarantee
• cost of product warranty.

In most cases, it makes sense to divide the costs of defects into the costs of identifying defects, the costs of eliminating defects, and the costs that resulted from defects.

The costs of conformity include the costs necessary to achieve compliance between the planned and existing quality; certification costs include all costs associated with documenting activities. These include costs for certification of quality management systems or costs for software that facilitates the distribution of documents throughout the enterprise. Control costs usually mean the costs of carrying out control activities before the start, during production and control of finished products, as well as the costs of all other quality control tools. This may also include external costs for providing guarantees, obtaining permits, etc. The costs of preventing defects include planning, performance research, supplier assessment, auditing, and staff training. This also includes production maintenance costs.

Practical examples of the use of quality control can be found in Almanac "Production Management"