Tree diagram. "seven tools" of quality management

When analyzing large amounts of data, we usually use the average value, less often the standard deviation, and even less often other processing methods. What causes this “self-restraint”? 🙂 Most likely, insufficient knowledge and experience in these matters. Where can a modern manager learn about statistical data processing methods? It is unlikely that he will remember the university statistics course. And was it included in the curriculum!?

My acquaintance with statistics, or more precisely with its use in business, began about 15 years ago, when I first read about quality management methods. Unfortunately, the seven basic tools “didn’t seem to me” the first time... I didn’t perceive them as a “guide to action.” Rather, I treated them as something transcendentally abstruse. And only gradually over the course of several years, repeatedly encountering the use of one or another method in the literature, as well as in connection with the emergence of practical problems, step by step, I began to understand the meaning of these tools and the scope of their application. Gradually, I began to use these methods in my practice, sometimes not even remembering that they are part of a coherent system.

The time has come to pay tribute to the original source - Japanese management, and also to show how seemingly book knowledge becomes a powerful tool for managing a real business.

Download the note in format, examples in format

Seven Basic Quality Control Tools Used to analytical problem solving, that is, in a situation where data is available, and in order to solve the problem, you need to analyze it.

1. Cause and effect diagram. This diagram is used to identify the process factors that influence the outcome. There are also names: “Ishikawa diagram” or “fishbone diagram”. IN classic version factors (reasons) are grouped into categories according to the “5M” principle:

Man (person) - reasons associated with the human factor; Machines (machines, equipment) - reasons related to equipment; Materials – reasons related to materials; Methods (methods, technology) - reasons related to the organization of business processes; Measurements - reasons associated with measurement methods.

Rice. 1. Ishikawa diagram. Sample.

It is clear that another relevant grouping can be used. For example, here is the “skeleton” we drew when analyzing the possibilities of reducing customer service time in a warehouse:

Rice. 2. Ishikawa diagram. Customer service time at the warehouse.

– a tool for collecting data and automatically organizing it to facilitate further use of the collected information.

Rice. 3. Check sheet. Example.

The advantage of checklists is that they can be used by employees who do not work with a computer. If the data for subsequent analysis is obtained by measurements directly at the workplace, checklists are very effective. It is clear that if the data for analysis is extracted from databases, checklists are not needed, and the data is immediately converted into a histogram, Pareto or scatter plot (see below).

In my practice, checklists have not found use, since the processes with which I deal are either completely related to the use of a computer, or are started by command from the computer, and the finish is recorded by the PC operator.

These charts rank issues by degree (frequency) of impact on the outcome. They got their name from the economist Vilfredo Pareto, who in one of his scientific works at the turn of the 19th and 20th centuries showed that in Italy 20% of households receive 80% of the income. The term “Pareto principle” was coined by the American quality management specialist Joseph Juran in the 40s of the 20th century. Pareto analysis is usually illustrated by a Pareto diagram, on which the causes of quality problems are plotted along the x-axis in descending order of their influence on the number of nonconformities (volume of defects), and along two ordinate axes: a) the number of nonconformities in pieces; b) the accumulated share (percentage) of the contribution to the total number of nonconformities. For example:

Rice. 4. Pareto diagram. Causes of overdue accounts receivable.

First of all, you should work with the reasons causing greatest number problems. In our example with the first three.

4. Histogram– a tool that allows you to visually evaluate the distribution of statistical data grouped by the frequency of falling into a certain (predetermined) interval. In the classic version, a histogram is used to identify problems by analyzing the shape of the scatter of values, the central value, its proximity to the nominal value, and the nature of the dispersion:

Rice. 5. Options for the location of the histogram in relation to the technological tolerance

Brief comments: a) everything is good: the average coincides with the nominal value, variability is within tolerances; b) the average should be shifted to match the nominal value; c) dispersion should be reduced; d) the mean should be shifted and the dispersion reduced; e) dispersion should be significantly reduced; f) two batches are mixed; should be divided into two histograms and analyzed; g) similar to the previous paragraph, only the situation is more critical; h) it is necessary to understand the reasons for such distribution; the “steep” left edge indicates some kind of action in relation to batches of parts; i) similar to the previous one.

Here are the histograms we have been building for several years to study customer service times in the warehouse:

Rice. 6. Histogram. Customer service time at the warehouse.

On the abscissa axis are 15-minute ranges of customer service time in the warehouse; The y-axis is the share of applications serviced in the allocated time range from the total number of applications for the year. The red dotted line shows the average service time throughout the year.

5. Scatter diagram(dispersion) is a tool that allows you to determine the type and strength of connection (correlation) between pairs of corresponding variables. These charts contain two sets of data plotted as dots. The relationship between these points shows the dependency between the corresponding data. In Excel, such a chart is of the “scatter” type. Here's an example of how I previously found the usefulness of scatter plots:

Rice. 7. Identification of correlation dependence based on a scatter diagram.

Here is an interesting example of using correlation analysis to manage the placement of goods in a warehouse:

The modern warehouse has very impressive dimensions. It can reach a depth of 100-150 meters (the distance from the loading gate to the back wall). It is clear that by placing goods with high turnover closer to the gate, you can save time moving around the warehouse. The figures above show the frequency of access to individual cells; on the left – for random placement of goods; on the right – for goods divided into ABC groups. The more intense the color, the more often the cell is accessed. It can be seen that without ABC distribution, access to cells is almost random; with ABC division of the nomenclature, zone boundaries can be observed. The left front of each figure faces the receiving area. Thus, in the situation depicted in Fig. b, the total path of storekeepers/equipment will be less than in Fig. A

6. Charts– a tool that allows you to analyze data across various sections. The forms and purposes of the analysis may dictate the use various types graphs. You can read more about this in Gene Zelazny's book "". Piece-by-piece comparisons of data are best demonstrated using a pie chart. A bar chart is best used to illustrate positional comparison. If component-wise and positional comparisons show relationships at a certain point in time, then temporal comparisons reflect the dynamics of change; Time comparisons are best illustrated with a histogram or graph.

For example, with these diagrams we analyze three parameters for each client at once: the dynamics of accounts receivable, overdue accounts receivable, and limits on the credit line:

Rice. 8. An example of using a graph to analyze data.

7. Control card– a tool that allows you to monitor the progress of a process and influence it, preventing deviations from the requirements presented to the process (or responding to deviations). There are two types of variations: natural, associated with the spread of values ​​around the nominal value inherent in the process; And special, the appearance of which can be explained by specific reasons. You can read more about this in the book by D. Wheeler and D. Chambers “. Business optimization using Shewhart control charts.” Control charts are used to identify special variations. The points corresponding to individual data, the line of average values ​​(μ), and the upper and lower control limits (μ ± 3σ) are plotted on the graph. If the points lie within the control limits, there is no need to react to deviations from the center line. If at least one point is outside the control limits, an analysis of the possible causes of the deviation is required. See, for example, "", "".

Using control charts to analyze the volume of accounts receivable:

Rice. 9. Control card. Natural causes of variation.

At week 27, the debt increased from $1.4 million to $2.6 million. However, no management action is required since the points were located within the control boundaries.

The following chart shows the average (by week) time for vehicles to take off:

Rice. 10. Control card. Special causes of variations.

It can be seen that, starting from the 19th week, the points go beyond the control limits. Process intervention is required to identify specific causes of variation.

I hope my examples will help you realize that the seven basic quality control tools can be a real aid to business process analysis.

They are presented according to the version given in the book by M. Imai “”. I have arranged these methods in the order that seems most logical to me.

OPTION 1:

Theory: Seven quality tools (graphical methods for assessing product quality)

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1. Seven simple quality tools. . . . . . . . . . . . . . . . . . . . . . . . . . .3

2. Cause-and-effect diagram (Ishikawa diagram). . . . 5

3. Checklists. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

4. Histograms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

5. Scatter diagrams. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

6. Pareto analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

7. Stratification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eleven

8. Control cards. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15

Task. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Literature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Introduction

IN modern world extremely important the problem of product quality arises. The well-being of any company and any supplier largely depends on its successful solution. Products more High Quality significantly increases the supplier’s chances in competing for sales markets and, most importantly, better satisfies the needs of consumers. Product quality is the most important indicator competitiveness of the enterprise.

Product quality comes from the process scientific research, design and technological developments, is ensured by good organization of production and, finally, it is maintained during operation or consumption. At all these stages, it is important to carry out timely control and obtain a reliable assessment of product quality.

To reduce costs and achieve a level of quality that satisfies the consumer, methods are needed that are aimed not at eliminating defects (inconsistencies) of finished products, but at preventing the causes of their occurrence during the production process.

The purpose of the work is to study seven tools in the field of product quality management in an enterprise. Research objectives: 1) Study of the stages of formation of quality control methods; 2) Study the essence of the seven quality tools. The object of the study is methods for studying the costs of product quality.

1. Seven simple quality tools

Control methods that have existed for a long time were reduced, as a rule, to analyzing defects through a complete inspection of manufactured products. In mass production, such control is very expensive. Calculations show that to ensure product quality through sorting, the control apparatus of enterprises must be five to six times greater than the number of production workers.

On the other hand, continuous control in mass production does not guarantee the absence of defective products in the accepted products. Experience shows that the inspector quickly gets tired, as a result of which some of the good products are mistaken for defective and vice versa. Practice also shows that where people are carried away by complete control, losses from defects increase sharply.

These reasons have forced production to switch to selective control.

Statistical methods make it possible to reasonably detect a process disorder even when two or three units of products selected for control turn out to be suitable, since they are highly sensitive to changes in the state of technological processes.

Over the years of hard work, specialists have isolated bit by bit from world experience such techniques and approaches that can be understood and effectively used without special training, and this was done in such a way as to ensure real achievements in solving the vast majority of problems that arise in real production.

One of the basic principles of quality management is to make decisions based on facts. This is most fully solved by the method of modeling processes, both production and management tools of mathematical statistics. However, modern statistical methods are quite difficult to understand and widely used. practical use without in-depth mathematical training of all participants in the process. By 1979, the Japanese Union of Scientists and Engineers (JUSE) had put together seven fairly easy-to-use visual methods for process analysis. Despite their simplicity, they maintain a connection with statistics and give professionals the opportunity to use their results and, if necessary, improve them.

These are the so-called seven simple methods:

1) Pareto chart;

2) Ishikawa scheme;

3) delamination (stratification);

4) control sheets;

5) histograms;

6) graphics (on a plane)

7) control charts (Shewhart).

Sometimes these methods are listed in a different order, which is not important, since they are supposed to be considered both as individual tools and as a system of methods, in which in each specific case the composition and structure of the working set of tools is supposed to be specifically determined.

The use of statistical methods is a very effective way to develop new technology and quality control of production processes. Many leading firms are committed to their extensive use, and some spend more than one hundred hours annually on in-house training in these techniques. Although knowledge of statistical methods is part of the normal education of an engineer, knowledge alone does not mean the ability to apply it. The ability to view events from a statistical perspective is more important than knowledge of the methods themselves. In addition, one must be able to honestly admit shortcomings and changes that have arisen and collect objective information.

2. Cause-and-effect diagram (Ishikawa diagram)

The 5M type diagram considers quality components such as “man”, “machine”, “material”, “method”, “control”, and in the 6M type diagram the “environment” component is added to them. In relation to the problem of qualimetric analysis being solved, for the “human” component it is necessary to determine factors related to the convenience and safety of performing operations; for the “machine” component - the relationship of the structural elements of the analyzed product with each other, associated with the implementation of this operation; for the “method” component - factors related to the productivity and accuracy of the operation performed; for the “material” component - factors associated with the absence of changes in the properties of the product materials during the execution of this operation; for the “control” component - factors associated with reliable recognition of errors in the process of performing an operation; for the “environment” component - factors associated with the impact of the environment on the product and the product on the environment.

Rice. 1 Example of Ishikawa diagram

3. Checklists

Checklists can be used for both qualitative and quantitative control.

Rice. 2 Checklists

4. Histograms

Histograms are one of the variants of a bar chart that displays the dependence of the frequency of the quality parameters of a product or process falling into a certain range of values ​​from these values.

The histogram is constructed as follows:

1. Determine the highest value of the quality indicator.

2. Determine the lowest value of the quality indicator.

3. Define the histogram range as the difference between the largest and smallest value.

4. Determine the number of histogram intervals. You can often use an approximate formula:

(number of intervals) = N (number of quality indicator values) For example, if the number of indicators = 50, the number of histogram intervals = 7.

5. Determine the length of the histogram interval = (histogram range) / (number of intervals).

6. We divide the histogram range into intervals.

7. Count the number of hits of results in each interval.

8. Determine the frequency of hits in the interval = (number of hits)/(total number of quality indicators)

9. Building a bar chart

5. Scatter plots

Scatter plots are graphs like the one shown below that show the correlation between two various factors.

Rice. 3 Scatter diagram: There is practically no relationship between quality indicators.

Rice. 4 Scatter diagram: There is a direct relationship between quality indicators

Rice. 5 Scatter diagram: There is an inverse relationship between quality indicators

6. Pareto Analysis

Pareto analysis gets its name from the Italian economist Vilfredo Pareto, who showed that most capital (80%) is in the hands of a small number of people (20%). Pareto developed logarithmic mathematical models that describe this heterogeneous distribution, and the mathematician M.Oa. Lorenz provided graphic illustrations.

The Pareto Rule is a “universal” principle that is applicable in many situations, and without a doubt - in solving quality problems. Joseph Juran noted the “universal” application of the Pareto principle to any group of causes that cause one or another consequence, with most of the consequences caused by a small number of causes. Pareto analysis ranks individual areas by significance or importance and calls for identifying and first eliminating those causes that cause the greatest number of problems (inconsistencies).

The seven quality control tools discussed are designed to analyze quantitative quality data. They make it possible to solve 95% of the problems of analysis and quality management in various fields using fairly simple, but at the same time, scientifically based methods. They use techniques mainly of mathematical statistics, are available to all participants in the production process and are used at almost all stages life cycle products.

However, when creating a new product, not all factors are of a numerical nature. There are factors that can only be verbal description. These factors account for approximately 5% of quality problems. These problems arise mainly in the field of managing processes, systems, teams, and when solving them along with statistical methods it is necessary to use the results of operational analysis, optimization theory, psychology, etc.

Union of Japanese Scientists and Engineers JUSE (Union of Japanese Scientists and Engineers) Based on these sciences, he has developed a very powerful and useful set of tools to facilitate the task of quality management when analyzing these factors. These tools are called the Seven New Quality Tools. These include:

• affinity diagram;
• diagram (graph) of relationships;
• tree (system) diagram (decision tree);
• matrix diagram or quality table;
• arrow diagram;
• diagram of the program implementation process (planning the implementation process);
• priority matrix (analysis of matrix data).

Affinity diagrams used to classify ideas (causes, indicators, problems, consequences, etc.) into groups united by a common character, the nature of these ideas.

To determine the causes of the problem working group using the brainstorming method to identify possible reasons, which are collected in the form of disparate data.

Systematize ideas that have a common focus into groups. This work is done without discussion. Names are not assigned to common characteristics.

If there are similarities between some groups, they can be combined into one larger group. At this stage, in the process of general discussion, the composition of the groups is agreed upon, some ideas are reformulated, combined or differentiated. Individual data can be transferred to other groups. We identify common feature for each group.

The composition of the data for each group is reviewed, the name of the groups and the final version of the generalizing feature are formed.

The clarity and ease of presentation of data provided by the affinity diagram is its indisputable advantage.

But the diagram also has a significant drawback - it is the subjectivity of the distribution of data according to related characteristics. This deficiency manifests itself most seriously during individual work. The brainstorming method and teamwork somewhat reduce subjectivity, but do not eliminate it.

Relationship diagram is intended for ranking related factors (conditions, causes, indicators, etc.) according to the strength of connection between them. The relationship diagram serves as a tool for identifying the most important, priority factors within each group. Conclusions are drawn on the basis of expert assessments during the brainstorming process.

• 1) write down each problem on a separate sheet of paper and attach the sheets of paper in a circle on the poster;
• 2) start from the top sheet and, moving clockwise, ask the question: “Is there a connection between these two events?” If there is, then ask: “Which event causes or causes another event to occur?”;
• 3) draw an arrow between two events, showing the direction of influence;
• 4) after identifying the relationships between all events, count the number of arrows emanating from each and entering each event.

The event with the largest number of outgoing arrows is the initial event. The team usually identifies two or three initial events that it must discuss to decide which one to focus on first. This takes into account various factors, for example, the organization’s limitations, resources, and experience.

Tree diagram. After identifying using a relationship diagram, the most important issues, characteristics, etc. using a tree diagram, they look for methods to solve these problems, ensure product characteristics, etc.

When searching for the causes of a problem, the “why-why” method is used. Members of the team that is solving the problem ask the question: “Why did it happen?” - and get a list of first-level reasons. Then the question “Why?” address each reason of the first level and receive a list of reasons of the second level, etc. The relationships between a problem (characteristic, etc.) and its causes at various levels are depicted in the form of a multi-stage tree structure. Schematic diagram such a diagram is shown in Fig. 8.22.

Rice. 8.22.

The benefits of a tree diagram are related to its clarity and ease of use and understanding. In addition, the tree diagram can easily be combined with other quality tools to complement them.

The disadvantages of this tool include the subjectivity of the arrangement of elements at a particular level of detail (especially if individual work is performed).

Matrix diagram allows you to visualize the relationships between various factors and the degree of their closeness. Problems in the field of quality and the reasons for their occurrence, problems and ways to eliminate them are analyzed, consumer properties products, their engineering characteristics, properties of the product and its components, characteristics of the organization’s performance and elements of the quality management system, etc.

The matrix diagram shown in Fig. 8.23 is the most common. It is called the /.-form, represents the relationship between two groups of factors, is widely used in structuring the quality function and therefore is called the quality table. Information about the degree of closeness of the relationship between various factors, presented using special symbols, allows you to more accurately model these relationships and more effectively manage various factors and processes.

Rice. 8.23.a b a 2,..., th, And b 2,..., b,- components of the studied objects A and B, which are characterized by different tightness of connections

Arrow diagram. After a preliminary analysis of the problem and ways to solve it, a work plan is drawn up to solve the problem, for example, to create a product. The plan must contain all stages of work and information about their duration. To facilitate the development and control of a work plan by increasing its visibility, an arrow diagram is used. An arrow chart can be in the form of either a Gantt chart or a network graph.

Figure 8.24 shows the order and timing of work on the construction of a turnkey house within 12 months, presented in the form of a Gantt chart.

The network graph for performing the same work is shown in Fig. 8.25. The numbers at the nodes of the graph correspond to the serial number of the operation shown in Fig. 8.24. In this case, the final operation corresponding to the “final inspection and delivery of the house” in Fig. 8.25 is divided into two operations: 11 - final inspection and 12 - handing over the house. The numbers under the arrows of the network graph correspond to the duration (number of months) of the operation, the number of which is indicated in the node of the graph from which the arrow comes.

Rice. 8.24.

Rice. 8.25.

Process Execution Planning DiagramPDPC (Process Decision Program Chart) used for planning, estimating the timing of complex processes in the field of scientific research, production of new products, solving management problems with many unknowns, when it is necessary to provide for various solution options, and the possibility of adjusting the work program. In this case, they first draw up a program and, if deviations from the planned points arise at the intermediate stages of its implementation, they focus on activities that bring the process into line with the program. When, during the execution of a program, an unforeseen situation arises that could not be taken into account in advance, it is necessary to draw up new program devoid of previous shortcomings.

Figure 8.26 shows an example PDPC- planning diagrams for the implementation of the process of selection and control of suppliers.

The benefits of a decision diagram are clear. With its help, you can see possible risks on the work execution plan and select one or another corrective action to reduce these risks. The disadvantages of this quality tool include high labor intensity if the plan has a significant number of tasks.

Priority Matrix is a tool that can be used to rank the importance of data and information obtained from brainstorming or matrix diagrams. Its application makes it possible to identify important data in situations where there are no objective criteria for determining its significance, or when people involved in the decision-making process have different opinions about the priority of data. The main purpose of a priority matrix is ​​to distribute different sets of elements in order of importance, as well as to establish the relative importance between elements through numerical values.

The priority matrix can be constructed in three ways. Construction options depend on the method of determining the criteria by which the priority of data is assessed - the analytical method, the method of determining criteria based on consensus and the matrix method.

The analytical method is used when the number of criteria is relatively small (no more than 6), it is necessary to obtain the full consent of all experts participating in the assessment, the number of experts does not exceed eight people, large losses are possible in case of an error in prioritization.

The method of determining criteria based on consensus is used when the number of experts is more than eight people, there is significant number criteria (from 6 to 15), there is a large number of ranked data (about 10-20 elements).

Rice. 8.26.

suppliers

The matrix method is used mainly when there is a strong relationship between the elements being ranked, and finding an element with greatest influence is critical for solving the problem.

The procedure for constructing the priority matrix for all three options is basically the same. The differences lie in determining the significance of the criteria.

The priority matrix is ​​constructed in the following order.

• 1. The main goal for the sake of which the priority matrix is ​​built is determined.
• 2. A team of experts is formed that will work on the task. Experts must understand the scope of the problem being solved and have an understanding of the methods teamwork(for example, about the brainstorming method, the Delphi method).
• 3. A list of possible solutions to the problem posed is compiled. The list can be compiled through the use of other quality tools, for example, brainstorming, Ishikawa diagram, etc.
• 4. The composition of the criteria is determined. Initially, it can be quite large. The priority matrix will include only part of these criteria, since in the future it will be reduced by selecting the most important and significant ones.
• 5. A weighting factor is assigned to each criterion. The weighting coefficient is assigned depending on the selected method.

For the analytical method:

• a rating scale is established for each criterion;
• For each numerical value of the scale, a definition of significance is given. In order to make differences in weighting coefficients more noticeable, a scale with numerical values ​​1-3-9 is usually used, where 1 is low significance, 3 is medium significance, 9 is high significance.

For the consensus method:

• a certain number of points is established that experts must distribute between the criteria. The number of points must be no less than the number of criteria;
• each expert distributes the assigned points between the criteria;
• The total number of points for each of the criteria is determined. This value will be the weighting coefficient of each criterion.

For matrix method:

• the criteria are arranged in the form of an /.-matrix;
• a scale is established for pairwise comparison of criteria (for example, “O” - criterion A is less significant than criterion B; “1” - criterion A and criterion B are equivalent; “2” - criterion A is more significant than criterion B);
• a pairwise comparison of all criteria is carried out;
• the weight coefficient of each criterion is determined (the weight coefficient is calculated as the sum of all values ​​in the matrix row).
• 6. The most significant criteria are selected. This can be done by discarding the criteria with lowest values weight coefficients. If the number of criteria is not large, then for further work all criteria can be saved.
• 7. A method is established for calculating the significance of each of the decisions of the priority matrix (defined in step 3) based on the selected criteria (defined in step 6).
• 8. Each decision is evaluated against each criterion.
• 9. The score is multiplied by the weight coefficient of the corresponding criterion. The obtained values ​​are summed up for each of the decisions, which gives a final assessment of the priority of the decisions. The final score, which the priority matrix contains, can be left as is, or converted into percentages.
• 10. The resulting list of solutions is sorted in order of priority. If necessary, the priority of decisions can be presented in the form of a Pareto chart.

Example 8.2

Build a priority matrix.

• 1. We determine the purpose of compiling a priority matrix: to reduce the number of defects in the product.
• 2. Forming a team of experts: for example, the team of experts will consist of three people. Each of them is familiar with the method of developing solutions based on brainstorming.
• 3. We make a list of possible solutions to the problem (generated by a team of experts):
  • change manufacturing technology;
  • conduct training for craftsmen;
  • change the product design.
• 4. We determine the composition of the criteria (the composition of the criteria for assessing the priority of decisions):
  • no more than 100 people/hour are required to implement the solution;
  • low cost of implementing the solution;
  • the number of personnel involved is no more than 50 people;
  • reduction in waste costs by at least 1.5 times.
• 5. Assign a weighting coefficient for each criterion. Let's consider the purpose of criteria for each of the three methods - analytical, consensus method and matrix method.

For analytical method

For the consensus method We establish that each expert can distribute 4 points between the criteria.

For the matrix method

Ending

• 6. Determine the most significant criteria: since the number of criteria selected for the example is only 4, we leave all the criteria.
• 7. Select a method for calculating the significance of each of the solutions proposed earlier (in step 3). To determine significance, we use a scale of 1-3-9, where 9 is the most significant decision, 3 is a significant decision, 1 is an insignificant decision.
• 8. We will assess the significance of each decision in relation to each criterion: to assess the significance of decisions, we will use the analytical method. The weighting coefficients of the criteria are determined in step 5.

Solution	Criterion
	Requires no more than 100 people/hour to implement the solution	Low price implementation solutions	The number of personnel involved is no more than 50 people	Reducing waste costs by at least 1.5 times
	Weighting factor 3	Weight factor 9	Weighting factor 1	Weight factor 9
Change manufacturing technology
Increase the number of control points
Conduct training for masters
Change product design

9. We determine the priority of each solution: the score of each solution is multiplied by the weighting coefficient of each criterion and the values ​​are summed up.

	Criterion
	Requires no more than 100 people/hour to implement the solution	price implementation	The number of personnel involved is no more than 50 people	Reducing waste costs by at least 1.5 times
	coefficient	coefficient	coefficient	coefficient
Change technology manufacturing
Increase the number of control points
Conduct education masters
Change design

• 10. We distribute solutions in order of priority:
  • conduct training of craftsmen - 118;
  • change manufacturing technology - 100;
  • increase the number of control points - 90;
  • change the design of the product - 72.

The priority matrix, compared to other ranking methods, makes it possible to more objectively assess the significance of the data and establish the value of this significance.

At the same time, the disadvantage of this quality tool is also obvious - it is very labor-intensive, especially when it is necessary to rank a large amount of data according to a large number of criteria.

These seven new tools are intended to complement other widely used statistical quality control methods. It is the joint use that is important known methods quality control and seven new quality control tools.

Test questions and assignments

• 1. Describe the features of statistical methods of quality control.
• 2. List the types of control charts for statistical regulation of technological processes.
• 3. What is the difference between control based on a quantitative characteristic and control based on an alternative characteristic?
• 4. Specify the procedure for constructing the control chart.
• 5. How is the data obtained on the control chart interpreted?
• 6. Draw an example of a control chart and explain the purpose of all the lines on the map.
• 7. How to manage technological process using control cards?
• 8. What is the reproducibility index and what does it reflect?
• 9. What kind of quality control is called selective?
• 10. Provide a diagram of defect levels. What is the difference?
• 11. What is a sampling plan?
• 12. What is the difference between the supplier’s risk and the consumer’s risk during selective product control?
• 13. Provide a diagram of one-stage and two-stage control plans. Explain the procedure for their implementation.
• 14. What are the operational characteristics of a sampling plan?
• 15. When and for what purpose are the “seven tools” of quality control used?
• 16. When and for what purpose are the seven new quality control tools used?
• 17. Describe the method of layering or stratification. For what purpose is it used in quality management?
• 18. What types of graphs do you know? For what purpose are they used in quality management?
• 19. Describe the Pareto diagram. For what purpose is it used in quality management?
• 20. Describe a cause-and-effect diagram. For what purpose is it used in quality management?
• 21. Describe the check sheet and histogram. For what purpose are they used in quality management?
• 22. Describe a scatter plot. For what purpose is it used in quality management?
• 23. Describe the diagrams used in quality management: affinity, relationships, tree, matrix, arrow, process diagram, priority matrix. For what purpose are they used in quality management?

Basic Concepts

The seven Japanese methods discussed above are designed for analyzing quantitative information. They allow you to solve up to 95% of quality problems. However, when creating, for example, a new product, not all factors are of a numerical nature. There are facts that can only be described verbally. They make up approximately 5% of problems in the field of managing processes and teams, and when solving them, along with statistical methods, it is necessary to use the results of operational analysis, psychology, and others.

Therefore, the Union of Japanese Scientists and Engineers developed 7 the latest tools , which allow us to solve these problems. These instruments were brought together and proposed by the Japan Union in 1979. These include:

1) Affinity diagram;

2) Dependency diagram;

3) System (tree) diagram;

4) Matrix diagram;

5) Arrow diagram;

6) Process evaluation planning diagram;

7) Analysis of matrix data.

Collection of input data for quality tools is usually carried out using the method brainstorming which is carried out with the help of specialists.

Scope of application of these methods: quality management, office work, education, training, etc.

Application of the "affinity diagram"

Affinity diagram– a tool that allows you to identify the main violations of the process by combining related oral data. It is a method of grouping together many similar or related ideas generated during a brainstorming session. The Japanese Union of Scientists and Engineers included the affinity diagram among the seven quality management methods in 1979.

The purpose of the method is to systematize and organize ideas, consumer requirements or opinions of group members expressed in connection with solving a problem. The affinity diagram provides general planning. It is a creative tool that helps to clarify unresolved problems by revealing previously invisible connections between individual pieces of information or ideas by collecting haphazard oral data from various sources and analyzing them according to the principle of mutual affinity (associative proximity).

Action plan:

1 Form a team of specialists who have knowledge of issues on the topic under discussion.

2 Formulate the question or problem in the form of a detailed sentence.

3 Conduct a brainstorming session related to the main reasons for the existence of the problem or answers to the questions posed.

4 Record all the statements on cards, group related data by area and assign headings to each group. Try to combine any of them under a common heading, creating a hierarchy.

The principles of creating an affinity diagram and identifying the main process violations in order to take measures to eliminate them are shown in Fig. 31. As can be seen from the figure, the affinity diagram is a creative means of organization large quantities oral data.

Figure 31 - Principle of constructing an affinity diagram

Additional Information:

The affinity diagram is used not to work with specific numerical data, but with verbal statements.

The affinity diagram should be used mainly when:

It is necessary to systematize a large number of information (various ideas, different points vision, etc.);

The answer or solution is not completely obvious to everyone;

Decision making requires consensus among team members (and perhaps other stakeholders) in order to work effectively.

Advantages of the method: p hides the relationship between different pieces of information.

The procedure of creating an affinity diagram allows team members to go beyond their usual thinking and helps to realize the creative potential of the team.

Disadvantages of the method: n In the presence of a large number of objects (starting from several dozen), the tools of creativity, which are based on human associative abilities, are inferior to the tools of logical analysis.

The Affinity Diagram is the first of the seven quality management techniques that helps develop a more precise understanding of a problem and identifies major process problems by collecting, summarizing, and analyzing a large amount of oral data based on the affinity relationships between each element.

9.2 Application of the “Interrelationship Diagram”

The relationship diagram is designed to rank related factors (conditions, causes, indicators, etc.) according to the strength of connection between them.

1) it is necessary to write down each problem on a separate sheet of paper and attach these sheets of paper in a circle;

2) you need to start from the top sheet and move clockwise, wondering if there is a connection between these two problems. If so, what event is the cause;

3) draw arrows between two events, showing the directions of influence;

5) the initial one is the one from which more arrows come out.

Example: Diagram of relationships to identify the causes of an increase in injuries at work In Fig. Figure 32 shows an example of a DV, reflecting the results of an analysis of the relationships between the causes of high injuries at work.

Figure 32 - Example of a relationship diagram

The Ishikawa diagram discussed earlier allows us to identify factors influencing any problem. The relationship diagram makes it possible to structure them based on their importance.

Thus, from this diagram it is clear that the main reasons for the increase in injuries during production are: lack of teamwork and insufficiently trained staff.

Seven Essential Quality Tools is the name given to a set of very simple graphical techniques that have been identified as most useful for solving simple, everyday quality issues. They're called main because even people with little or no statistical training will be able to understand these principles and apply them to their daily work.

I have often seen that even highly qualified personnel ignore the idea of ​​using modern instruments qualities such as experimental design, hypothesis testing, or multivariate analysis. Although it would be useful for most professionals to know that majority quality issues can be solved using these seven essential quality tools.

The purpose of this article is to review these basic tools and their effective use. Getting the best results with any of these tools does not require proof; The quality specialist must provide complete, objective and sufficient information.

Tool #1: Ishikawa diagrams

(also called " fish skeleton" or " cause-and-effect diagrams") are cause-and-effect diagrams that show the root cause(s) of a particular event. A common way to build a truly informative fishbone is to use the 5 Whys method and a cause-and-effect diagram together.

1. People - Personnel involved in the process; stakeholders, etc.
2. Methods - Processes for performing tasks and specific requirements for performing them, such as policies, procedures, rules, regulations and laws
3. Machinery - Any equipment, computers, tools, etc. needed to perform the job
4. Materials - Raw materials, parts, pens, paper, etc. used for production final product
5. Indicators - Data obtained from a process that is used to evaluate its quality
6. Environment- Conditions such as location, time, temperature and culture in which this process is carried out

Tool #2: Checklist

It is a structured, prepared form for collecting and analyzing data. This is a versatile tool that can be adapted for a wide variety of purposes. The data collected may be quantitative or qualitative. When the information is quantitative, the checklist is called accounting sheet.

The defining characteristic of a checklist is that data is entered into it in the form of marks (“checkmarks”). A typical check sheet is divided into columns, and the marks made in different columns have different meanings. The data is read based on the location and number of marks on the sheet. Checklists typically use a “header” that answers five questions: Who? What? Where? When? Why? Develop operational definitions for each of the questions.

1. Who filled out the checklist?
2. What was collected (what each mark, lot identification number, or number of items in the lot represents)
3. Where did the data collection take place (equipment, premises, tools)
4. When the data was collected (hour, shift, day of week)
5. Why this data was collected

Tool #3:

Is a display statistical information, which is represented by rectangles to show the frequency of data items in successive numerical intervals of the same size. In the most common form of a histogram, the independent variable is plotted on the horizontal axis and the dependent variable is plotted along the horizontal axis. vertical axis.

The main purpose of a histogram is to clarify the data presented. It is a useful tool for plotting processed data into areas or bars of a histogram to establish the frequency of certain events or categories of data. These histograms can help reflect the highest frequency. Typical applications of root cause analysis histograms include presenting data to determine the dominant cause; understanding the distribution of manifestations of various problems, causes, consequences, etc. A Pareto chart (explained later in the article) is a special type of histogram.

Tool #4:

Is an important tool and solution. Since organizational resources are limited, it is important for process owners and stakeholders to understand the root causes of errors, defects, etc. Pareto excels at representing this mechanism by clearly ranking the root causes of a defect. The diagram is also known as the 80:20 principle.

A chart, named after economist and political scientist Vilfredo Pareto, is a type of graph that contains bars and line graph, where individual values ​​are represented in descending order by columns and the accumulated sum is represented by a line. The left vertical axis usually represents the frequency of occurrences. The right vertical axis is the total percentage of the total number of manifestations. Since the causes are arranged in descending order of their importance, the cumulative function is concave. As an example of the above, in order to reduce the number of tardiness by 78%, it is enough to eliminate the first three reasons.

Tool #5: Scatter plot or scatter plot

Often used to identify potential relationships between two variables, where one may be considered an explanatory variable and the other a dependent variable. This gives a good visual picture of the relationship between two variables, and helps in analyzing the correlation coefficient and regression model. The data is displayed as a set of points, each of which has the value of one variable that defines the position on the horizontal axis and the value of a second variable that defines the position on the vertical axis.

A scatter plot is used when there is a variable that is under the control of the experimenter. If there is a parameter that systematically increases and/or decreases when influenced by another, it is called control parameter or independent variable and is usually plotted along the horizontal axis. The manipulated or dependent variable is usually plotted along the vertical axis. If there is no dependent variable, or the variable can be plotted on any of the axes or on a scatterplot, it will only show the degree of correlation (not the cause-and-effect relationship) between the two variables.

Tool #6:

It is a method of sampling the population. In statistical surveys, when the population groups in the population are different, it is advisable to sample each group (stratum) separately. Stratification is the process of dividing members of a society into homogeneous subgroups before sampling.

The strata must be mutually exclusive: each population unit must be assigned to only one stratum. The strata must be exhaustive: no population unit can be excluded. A simple random sample or a systematic sample is then taken within each stratum.

This often improves the representativeness of the sample by reducing sampling error. It can produce a weighted average that has less variability than the arithmetic mean of a simple random sample of the population. I often tell the groups I supervise that proper selection procedures are more important than just having a sufficient sample size!!

Tool #7: Control charts, also known as Shewhart charts or process behavior charts

It is a special type of timing diagram that allows significant change differentiate due to the natural variability of the process.

If control chart analysis shows that the process is under control (ie, stable, changing only due to reasons inherent to the process), then no corrections or changes to the process control parameter are required or desired. Additionally, data from this process can be used to predict future process performance.

If a map shows that an observed process is out of control, analysis of the map can help identify sources of variation that can then be addressed to bring the process back under control.

A control chart can be seen as part of an objective and disciplined approach that helps right decisions regarding process control, including whether process control parameters need to be changed. Process parameters should not be adjusted for a process that is under control, as this will reduce process performance. A process that is stable but is operating outside of a given range (the scrap rate, for example, may be statistically controllable but above a given norm) must be improved through focused efforts to understand the causes of current performance and fundamentally improve the process.

When I manage simple projects ( Six Sigma) (usually called a yellow belt project), where the issues are not complex and the project team consists of people with 3 to 5 years of experience in the process, I strongly advocate the use of these simple tools to resolve process issues.

As a rule of thumb, any process demonstrating 1-2% repeatability standard deviations,can be improved by simple analysis using these tools. Only when process reproducibility is greater than 2.5 - 3% standard deviation should medium to advanced tools be used to identify and resolve process issues. I also recommend to anyone initial course Six Sigma education and training use the seven quality control tools to create fertile ground for the development of green and black belts within the organization.

Material prepared by Andrey Garin
based on materials from foreign publications
http://www.site/