Do you have money, but don’t need knowledge? The usual practice of the dominant management style is while there is money, and when there is no money, knowledge will no longer help.
Material prepared by: Scientific Director of the AQT Center Sergey P. Grigoryev
Free access to articles does not in any way diminish the value of the materials contained in them.
Preamble
The material presented below does not contain numbers (for obvious reasons), but in terms of the identified shortcomings presented, they can be attributed to many, at first glance, successful enterprises.
Company "N" is considered a very successful enterprise; it is constantly visited by delegations on excursions, from one to three visits a day.
"A typical situation: the management of a company is eager to improve quality and productivity, but does not understand at all how to do this. In search of enlightenment, it visits other companies, which at first glance seem prosperous. There, these managers are received with open arms, and an exchange of ideas begins. They (visitors ) learn how the company operates, some of it may accidentally coincide with the 14 points. Without using the guidelines, both are adrift. Neither company knows which procedures are correct or incorrect and why. "The question is why it is successful and whether it could be made even more successful. One can only hope that visitors will enjoy the tour. It is not their fault, but their problem."
It should be noted that I have never seen such a conspicuous order in any enterprise, from production facilities to the shelf for safety glasses in the workers’ restroom. As it turned out later, this was the only thing that could be done with only desire and financial resources.
The first day of my visit began with attending a weekly meeting with the heads of all departments of the company. From the reports it was clear that managers do not work together to optimize the company’s activities as a whole and everyone is busy with the goals (KPIs) of their department, in fact, sub-optimization, which only brings harm to the overall goals of the company.
“When there is no agreement among the comrades, their business will not go well, and nothing will come out of it, only torment. One day Swan, Cancer, and Pike took on a load of luggage, and together the three of them harnessed themselves to it; They climbed out of their skins, but the cart still doesn’t move! The luggage would seem light to them: Yes, the Swan is rushing into the clouds, the Cancer is moving back, and the Pike is pulling into the water. It’s not for us to judge who is to blame, who is right;
At the meeting, the speakers compared the plan with the fact for the reporting period and in relation to the previous reporting period in the form of tables with numbers; for some, the numbers were supplemented with graphs with the corresponding columns. In rare cases, the fact reached the planned indicators, which needlessly embarrassed the speakers. It was clear that nature of variability and the use of statistical methods to analyze performance indicators and predict their future behavior, no one at the enterprise knows.
Production and quality control
To control quality in production “in tolerance/out of tolerance”, gauges are used, for example, plugs and staples. For turning production, this is quite normal practice. But such a division of products into good and defective deprives workers and company management of the most important information hidden behind the discrete data necessary for the constant and effective improvement of parts manufacturing processes.
By the way, black and white TVs were quite in demand at one time.
Figure 1. Information that can be obtained when using gauges in quality control. The percentages in the figure are not related to this company, but are shown for clarity.
The failure to use statistical methods (Shewhart control charts) for research, even the discrete data available in the company, indicates the absence of systematic activities in the company, pursuing the goal of continuous improvement of production processes.
Of course, the company keeps statistics on the percentage of defective parts and displays graphs for everyone to see, but no one knows what to do with this data next.
A special source of pride for the company’s management is the “Diamond” system, which was thoughtlessly copied from the Instrum-Rand company, the brief essence of which is the “voluntary” surrender by workers of defective products they produce, determined, among other things, with respect to tolerance limits. When an employee voluntarily surrenders a defective product, he is not punished, but is fined for failure to surrender.
The word “punishment” is taken from demonstration materials about the “Diamond” system of this company and, in my opinion, sounds derogatory towards the current employees of the enterprise. The possibility of unfair “punishment” only aggravates the situation (more on this below).
“External motivation is the mentality of those who call for “zero defects.” A person begins to fight to preserve what he has. He tries to avoid punishment. But all this is humiliating and leads to personality degradation: it’s just daily wages for daily labor. But Pay beyond a certain level does not provide motivation. Pay certainly does not provide intrinsic motivation, and it is certainly not as powerful an extrinsic motivator as most people tend to think. We have to go back to the individual, to give the person the satisfaction of being what he is. does."
It is important to note that the author of the “Brilliant” system, General Director of the Instrum-Rand company, Vadim Sorokin (1993), now President and General Director of the GAZ Group, where the “Brilliant” system has not been implemented to this day, did not lay down Brilliant imposes fines for failure to deliver defective parts. But managers of Russian enterprises cannot accept this system without mandatory fines for failure to submit. Their paradigms stand guard over a vicious management style. Otherwise management would have to get on with their job.
"Copying is risky. You need to take on something if you are already savvy in theory. Americans are great imitators (for example, quality circles, kanban or just-in-time). However, the Japanese study theory first and only then start improving. American Management, not understanding its role in the work of "quality circles", tries to copy them, only to discover after a while that it is all false. "Quality circles" that cooperate and interact with management will work well everywhere.
Even the original “Diamond” system has its significant drawbacks, namely:
- tolerance boundaries are used as boundaries of actions in relation to the process, which maximizes business losses from committing errors of the first and second kind ;
- there is not a word about checking for stability, bias and error of the measurement systems of each operator and controllers;
- there is not a word about checking the stability of production processes that produce defective parts;
- payment is provided for the provision of identified defective products, as if operators working in a cooperative environment would not do this for free.
“The more you pay people to do something, the more their interest shifts from the activity they have to do to the reward.”
Next, all those involved in the enterprise begin to work with the delivered products that are defective within the tolerance limits in order to identify the causes of the defects. The authors of the "Diamond" system, the management of this company and those involved in the analysis of defective products lose sight of the fact that defective and defect-free products relative to the tolerance limits can be a natural output of a stable process, i.e. be completely homogeneous - produced by one system, for which The management of the enterprise is responsible, not the workers. With the same “success”, in this case, it would be possible to study defect-free products produced by this process.
"Any two numbers that are not the same are considered different. Unfortunately, this is true when it comes to arithmetic, but it is not true when it comes to interpreting data. In this world, two different numbers may well represent the same thing."
If a process producing products defective relative to the tolerance limits is in a statistically uncontrollable state, such actions may have some positive impact on the output of the process, but are always significantly less effective (effort/result), since there is a high probability of missing special causes of uncontrollability , which could be detected by the Shewhart control chart at points located inside the tolerance zone, which leads to "night blindness" of observers - failure to use additional opportunities to improve the process.
Using tolerance limits to take action against participants in a process is ignorant. If the process is in a statistically controlled state, such measures will interfere with the operation of a stable system, leading only to deterioration. See Edwards Deming's experiment for an explanation of this phenomenon. "Funnel and Target" and an article by Donald Wheeler "Right and Wrong Ways to Use Tolerance Fields." .
Explanation. Tolerance limits are consumer requirements (external and internal), often not related to the capabilities of the process. The process boundaries on the Shewhart control chart are the real capabilities of the process. Of course, it is desirable that the capabilities of the process meet the requirements for it and even exceed them. But compliance with requirements can be achieved in two ways or a combination of them: 1. Improve process capabilities; 2. Change the requirements for the process (tolerances), for example, agree with the consumer to expand the tolerance limits.
Video 1. What needs to be done to improve processes?
Symbols of elements in the video: НГД and ВГД - lower and upper tolerance limits, respectively (Eng, LSL and USL); m0 - nominal tolerance field; НГП and ВГП - lower and upper natural boundaries of the process (English LNPL and UNPL); CL - central line of the process (average of the process).
At the same time, if your processes operate wider than the tolerance limits, you will have to sort products into defective and non-defective before shipping them to the consumer, making 100% control, with the ensuing consequences described in the article by Donald Wheeler: “Is the product on specification actually compliant?”
Note that a statistically stable (stable) state of a process producing defective items would indicate the best it can do in the system in which it operates. In this case, only systemic changes, for which the company's management, and not workers, are responsible, could significantly change the situation.
"No more than 6% of all problems (or opportunities for improvement) in organizations are associated with special causes of variation; thus, they are possibly (but not necessarily!) in the field of activity of ordinary employees. In this case, top managers account for at least 94 % of all potential improvements to the system in which their employees work.
No control and no level of professional skill of workers will be able to overcome the fundamental defects of the system."
Conversely, processes that are in a statistically unstable (unstable) state operate at maximum costs under the influence of special reasons that must be identified at the shop level and eliminated. It is important to understand the signs (rules) that distinguish a stable state of processes from an unstable one, and these are not tolerance limits! See the explanation in the article: Nature of Variability .
“Statistical control can be achieved by hunting for each specific cause and identifying it whenever a point falls outside the control limits and taking appropriate action.”
Refusal to record the values of controlled measurements of all products (good and defective) selected for control, in the form of continuous values, hides information about the statistical state of processes and statistics of data distribution regarding the nominal value and tolerance limits. Such records are kept only for defective products in defect reports, where this data is buried.
Recording all measurements selected from the general flow of products and maintaining Shewhart control charts directly by machine operators would allow:
- find out the type of reasons causing the observed variations (general system or special reasons) as soon as the part selected for inspection is measured by the operator;
- take evidence-based methods to assist operators while minimizing the risk of committing errors of the first and second kind .
For example, if the quality indicators demonstrate a reasonable degree of statistical control, and the distribution histogram is shifted to the right or left relative to the tolerance fields, possible reasons for such a shift could be: a bias in the measurement system; machine settings by the operator or technical condition of the machine. In this case, there is a subject for research, and not for interrogating the worker about the reasons for the marriage. Making a decision to change quality management methods at an enterprise is the responsibility of management.
The two parts of the slide below show the same percentage of defective products (graphs provided by Donald Wheeler). Where is more useful information for improvements?
Figure 2. Information that can be obtained by using gauges (black and white picture on the left) and statistical methods (color picture on the right) in quality control in production.
"There is no substitute for knowledge. But the prospect of using knowledge is scary."
The metrological service of the enterprise has set up and equipped an excellent stand for training workers in the use of measuring instruments. There is a separate employee for training. But no one is interested in the fact that no one teaches students how to correctly record the obtained measurements. Some operators in quality reports, when recording measurement results, separate the integer and decimal parts of the number with a “-” sign. Thus, the “-” sign in reports goes through all stages from the controller at the production site to the department where all data is entered into a single database on a computer. No one who sees these entries in the reports pays attention to this.
All equipment of this enterprise contains signs with the slogans “Three NOTs” (“Do not accept defects” - “Do not make defects” - “Do not pass on defects”). These slogans are very popular in Russian industry today and are posted everywhere. As if the workers, if only they wanted, could eliminate the defects that are the resulting output of the processes that the enterprise management designed.
“Slogans and posters are addressed to the wrong people!”
I’ll make a small digression from this rapid examination and tell you a story.
At one manufacturing plant that I led, we were faced with the situation of identifying a new defect in the finished product, namely a visible step at the welded joint of two parts of the structure, which significantly affected the appearance and strength characteristics of the finished product.
Together with the management of the enterprise, we followed the entire chain of production operations “upstream” to the design department. As it turned out, a problem that diligent workers could not cope with was “not worth a damn.” The fact is that the designer-designer, working in CAD, marked the location of the weld with entry into the zone in which the weld should not be, perhaps believing that “it will be stronger.” The designer did not even know that he had made a mistake during the design. Next, the welder, a professional of the highest class, who has an informal high authority in the work environment, responsibly did his job and transferred the part to the resistance spot welding station, where the two parts of the product are connected.
I would like to emphasize that the work performed by the welder was not a defect by definition and was carried out in accordance with the technical specifications. The weld made in the previous operation created an unacceptable gap at the spot welding site. The resistance welder also did his job responsibly, while making the resistance welding an unreliable connection with the appearance of a noticeable step at the junction of the two parts. Here, the resistance welder could have wondered about the height of the step in the joint, but no one gave him an operational definition of the observed defect, such a definition simply did not exist and he knew that he was doing his best, and a more qualified and authoritative worker was working on the previous operation. Next, the structure was transferred to the grinding and polishing station, where a diligent worker had to “reduce to nothing” the place where the parts were joined (clean and polish). The step that arose during previous operations caused great difficulties for this worker, but no one gave him an operational definition of the observed defect.
Any doubts the worker had during the stripping operation disappeared when he realized that the previous operations had been carried out by more experienced workers. He needed more time for the stripping operation and made it impossible to completely hide the joining line, but his job was to reduce the step to nothing, and not to understand the reasons for its occurrence in previous operations. No one gave him an operational definition of what “no” is. So this assembly unit ended up in the assembly of finished products.
We gathered all the participants in the chain of the above production operations in the workshop and demonstrated what difficulties their work causes at each subsequent operation until the assembly of the finished product. The fact is that they were all responsible people doing their job, they simply did not know about the emerging problem with their products in the following operations. It never occurred to anyone to approach a worker at a previous operation and talk about the problems he was facing. Naturally, this defect did not appear again. Management should have just minded their own business.
Would a poster with “Three NOTs” hanging here help?! And in this and in any other cases, posters with calls to improve quality and increase productivity addressed to employees working in the system designed by the enterprise management are useless and cause justified indignation. It is better to hang posters explaining what management is doing to improve quality, increase productivity and optimize the activities of all departments according to the company's goals as a whole.
See how it all ends E. Deming's experiments with red beads , in which the master makes attempts to motivate workers to work without defects and take a responsible attitude towards their duties using monetary rewards, threats, persuasion and slogans.
"No more than 6% of all problems (or opportunities for improvement) in organizations are associated with special causes of variation; thus, they are possibly (but not necessarily!) in the field of activity of ordinary employees. In this case, top managers account for at least 94 % of all potential improvements to the system in which their employees work.
No control and no level of professional skill of workers will be able to overcome the fundamental defects of the system."
Further based on the results of the express examination
External manifestation of concern for the safety of enterprise employees and visitors (safety glasses, hard hats) - everywhere except in those places where it is really dangerous. In front of the entrance from the courtyard to the office building where the process optimization department is located, icicles weighing up to 3 kg periodically fly from the roof. There are not even signs of fencing off the danger zone or warning of danger. After my comment to the head of the department about this risk in the morning of the working day, nothing changed until the end of the day.
In a written report on defective products produced, the worker explains the cause of the defect by his fatigue after a sleepless night. No one even bothered that a defective product was the least that could happen to a tired lathe operator.
The result of the survey is presented to the head of the enterprise at a presentation in the form of a report on unused reserves and identified problems with the proposal of measures aimed at continuous improvement and increasing the competitiveness of the enterprise. Apparently, no one except the head of the company found out about this report.