Introduction
All must be aware of the often quoted famous saying of H.G.Wells that “Statistical thinking will one day be as necessary for effective citizenship as the ability to read and write”.
This was well understood by USA. They use very effectively statistical system in all aspects of life, particularly in commercial area.
Today, SQC system is considered to be the most essential requirement for industry, whether it is manufacturing or service. Dr W A Shewhart, Dr W E Deming, Dr J M Juran, Dr Taguchi and Dr K Ishikawa insist upon the effective use of this system to achieve excellence in quality. Knowledge and practice of SQC system is therefore equally important for Quality Circle functions and Quality Improvement Team activities.
Use of SQC methods
These techniques started as a sporadic use in Japan. After the Second World War, JUSE with the help of Mr. W.E.Deming and Dr. J.M.Juran carried out intensive propagation and brought them to regular use in the industries. Dr K. Ishikawa who worked with them divided statistical methods into three categories.
1. Elementary Statistical Methods (7-tools) consisting of:
i) Check-sheet (for data collection)
ii) Stratification
iii) Graph and Control Chart
iv) Pareto Chart
v) Cause and Effect Diagram
vi) Histogram
vii) Scatter Diagram
He considers them as seven indispensable tools to be used by everyone i.e. Company Managing Directors, Company Directors, Middle Management, Foremen and line workers. It can be used anywhere. According to him 95% of all the problems in a company can be solved by these tools. In Japan these tools are used effectively by top management to line workers.
Along with these tools, facilitators and workers must also be trained in the following basic concepts for efficient functioning of Quality Circles.
a. The concept of quality.
b. Principles and implementation concerning ‘management and improvement’.
c. Statistical way of thinking.
2. Intermediate Statistical Methods consisting of:
i. Theory of sample surveys.
ii. Statistical sampling inspection
iii. Various methods of making statistical estimates and tests.
iv. Methods of utilising sensory tests.
v. Methods of design of experiments.
These methods are taught to engineers and quality control personnel effectively in Japan.
3. Advanced Statistical Methods constitute a big list but the most important ones are:
i. Advanced methods of design of experiments
ii. Multivariate analysis.
ii. Various methods of operation research.
They are used by technicians and engineers for optimization of processes and securing reduction in costs.
The use of all these three methods with computers wherever necessary, is reported to have played an important role in the quality development in Japanese industry. (A Blackbelt under Six Sigma need to know all these methods)
Statistical methods and technological advance
The use of statistical methods including the most sophisticated methods has become deeply rooted in Japan. However, one must not forget the utility of simple seven tools. Unless a person masters these seven tools he cannot use the more sophisticated ones. As Dr K Ishikawa says “It was due to the use of this that the quality level has risen, reliability has risen, and cost has fallen. The key has been the dogged use of process analysis and quality analysis without fanfare for a long period of time. This has brought about improvement in technology. Some contend that engineering technology enhances technology and management technology maintains it. I do not subscribe to that claim. I cannot see any difference between engineering technology and management technology. The so-called control technique is part of proper technique. One must utilize all the technology at his disposal to strive towards advancing quality and efficiency. After the Second World War, Japan imported many new technologies from the West. Nowadays, Japan can export her technologies to the West as well. This is in large measure a result of the introduction of statistical quality control and the use of statistical analysis, process analysis and quality analysis.”
The late Prof. P.C. Mahalanobis, founder Director of Indian Statistical Institute, foresaw long ago the role of statistics in national development when he said, “Statistics is the key technology of this century”. Keeping in mind what is stated by Prof. Mahalanobis and Dr. K Ishikawa, we strongly advocate knowledge and effective use of basic elementary statistical methods. After careful analysis, we have separated graphs and control charts as two separate tools and added to that two more tools viz. Flow diagram and Brainstorming. We call them Problem Solving Techniques. It is absolutely necessary that everyone from the Chairman, Managing Director to grassroots level employees should learn and effectively make use of the ten problem solving tools. We are listing below the tools along with their use.
Overview of the 10 Tools
1. Flow Diagram: To enable understanding of the process/flow and
locate a problem.
2. Brainstorming: For generation of ideas in problem listing; listing of causes in problem analysis and for finding solutions.
3. Data Collection: To understand the magnitude of the problem and understand the problem. It is the foundation for statistical analysis.
4. Graphs: Presentation of large amount of data in a cohesive manner in the pictorial form to enable better understanding of the data and the problem and also to comprehend the trend at a glance.
5. Stratification: To segregate data according to contributing sources (Suppliers, machines, operators etc.).
6. Cause & Effect Diagram: To map out all probable causes and relate the logical linking of causes to the problem to help narrow down to the cause or causes.
7. Pareto Chart: For identification/selection of major problem or area for improvement or control; to differentiate between vital and trivial problems. 8. Scatter Diagram: For examining relationship between two variables; nature and strength of relationship between process factors and product quality. In a way, it is also a cause and effect approach between two variables.
9. Histogram: For study of process variation and assess process capability. This is an off-line quality control method.
10. Control Charts: For maintaining running control on a process. It is a tool for on-line quality control.
Procedures to be followed by Quality Circles in making use of these techniques.
For nearly two decades we are observing the case studies presented by Quality Circles in our various conventions. While explaining the procedure here the deficiencies observed in them are taken into account. Quality Circles meet regularly and periodically. The meeting is normally once a week for one hour.
The following steps outline the procedure along with reasons.
Step 1 – Identification of Problems
The identification of the problem is to be done by using structured method of‘ Brainstorming’ technique. This technique is to be used for also finding out the causes for a problem, finding solution(s) and possible “resistance to the solutions” and how to overcome those. This is an important technique and generally the members, Facilitators and the organisation pay not much attention to this technique. If well carried out, a group of five or six members can identify more than hundred problems at any time. With more members, the problems identified could be much more.
The brain is to be raked repeatedly at the various stages of problem solving to improve the logical thinking and also to bring out the creative ability of each one.
After identification of the problems next step is selection of a problem for finding solution. A logical step at this stage is discussing with the Facilitator and the department about the list of problems prepared.
Some of them may be irrelevant, some may call for a small discussion and/or clarification for finding a solution and eliminating them. They have to be done. Rest of the problems in the list will call for a detailied analysis. How it is to be done is explained in the next step.
Step 2 – Selection of the Problem
The ideal way to select a problem is based on data. This is the approach senior managers are expected to follow when they work in Quality Improvement Team. There are many problems encountered in this method when workers take it up. In fact many of them have yet to learn what is data, how to collect and make use of it.
So ‘A’, ‘B’ and ‘C’ categorization and rating methods are suggested for them.
“A” Category: Minimum involvement of other department in solving them
“B” Category: Involvement of other departments is a necessity
“C” Category: Management sanction may be needed in implementing the solution
Once the listing is done the selection process starts. We want Quality Circle to take up ‘A’ category problem.
Generally the number of problems in this category will be high. First categorize them into groups. The grouping can be done based on certain basis. For example problems related to
i) safety
ii) environmental pollution
iii) working condition
iv) cleanliness
v) scrap or wastage
vi) blemishes leading to rework or reprocess etc.
Take up a group, which the team and the Facilitator consider as important. Here too you will have a list of problems.
Ideal way to select a problem will be based on data.
But generally workmen have no access to the data and many of them are not familiar with that. Since we need involvement of all, adopting a method, which will involve all is better. That is why ‘Rating Method’ is suggested for Quality Circles. Using this method select a problem for finding solution.
Workmen use different basis for selection of a problem in some organisation. It does not matter. All what we want is that they should understand and have a systematic approach.
Here we would like to mention about problems encountered by some of the organisations in this aspect.
The reason why we are suggesting Quality Circles to take up ‘A’ category problem for finding solution is to make them learn gradually. Normally almost all ‘A’ category problems can be solved by simple discussion with minimum data. As we mentioned earlier ‘A’ category problems will be maximum and also very simple to find solution and all of them can be solved within a year.
But we are informed by some of the organisations that it is not happening in that way. Since Quality Circles are interested in preparing a case study for presentation they do not find ‘A’ category problems suitable for that and hence take a bigger problem for finding solution. Since it suits the organisation also, they do not object to that.
In some cases Quality Circles do take up an ‘A’ category problem for finding solution and solve the problem in a laborious way taking lot of time.
In a Quality Circle convention held in a reputed organisation recently the CEO pointed out that almost all Quality Circles have identified more than 60-80 problems. But they have solved only two or three in a year. He wondered that how many years it will take to get rid of all the problems. He pointed out that it is not a satisfactory situation. We agree with him. The issue need to be taken seriously by the department head concerned and also by the steering committee.
We suggest that the steering committee and the department heads should ensure that all minor problems belonging to ‘A’ and ‘B’ categories are found solution within a stipulated time. Department head should take it as his responsibility to compile all the problems of department identified by the Quality Circles. He should ensure that besides Quality Circles others are also involved in solving the department problems. It should not be left as the responsibility of Quality Circles alone.
Once a problem is selected the next step is estimating the time needed for finding the solution i.e. how many meetings need to be conducted for finding the solution. For many problems workmen will be able to arrive at a solution by just discussing it. But we have a dual purpose. Most important is development of people and second one is solution to the problem. In fact, as we said earlier, the whole exercise is undertaken to educate the persons involved. So there is no point in rushing with the process. Another important point is, it is found that many times when a group carries out a systematic analysis more clarity about the situation and better understanding of the process results. This leads to problem prevention.
So make a Milestone chart or Gantt chart for it. What is the difference between a Milestone Chart and a Gantt Chart? They were made by two different persons for guidance when a job calls for planning. They are very useful when the planning need is simple and straight forward. Where multiple activities are involved CPM (Critical Path Method) or PERT (Programme Evaluation Review Technique) is used. Now-a-days these two are replaced by Arrow Diagram.
How to make a ‘Milestone Chart’?
Once again our observation is that this is not done sincerely by most of the Circles. Just filling up the format is not sufficient. Milestone chart is a planning activity i.e. how the problem solving is to be carried out and how much time the circle expects to spend on each activity. It is an estimation and has to be done before the problem is taken up for solving. The proper way is to take a sheet of white paper and write down all the steps involved in problem solving one by one, in the order it is to be taken up and the time each activity is estimated to take up. Quality Circle members should write all the steps involved in problem solving one by one in a white sheet.
Quality Circles have already identified the problems related to their area and also have selected a problem for finding solution. Next step is ‘Problem definition’ and ‘analysis of problem. These should be discussed after looking into past data, study on the shop floor, discussions with others, both within the department and outside. The time required in terms of number of meetings etc. should be estimated. This should be depicted in the Milestone chart. Similarly estimation of the number of meetings it would take for all the other activities should be depicted on the milestone chart after proper discussion and estimation.
Later-on when they start proceeding with the efforts to solve that problem, after completion of each step before proceeding to the next step they should compare the actual time taken with the estimated time. They should mention in another paper step by step the actual time taken and reasons for variation wherever occurred. Along with the final Chart which shows estimated as well as actual in two colours the estimation process papers also have to be shown.
In the initial stage they are bound to make errors in the estimation. They will start learning by making the necessary corrections and over a period will be able to make a proper estimation.
Learning how to estimate the time needed and how to proceed is an important aspect for everyone, more so for people who do not have money and manpower to carry out their errands. Unnecessary delays and losses due to not planning their activities can be avoided only by learning this system. Our request to the Circles and the organisations is not to take this in a casual way and lose the benefit they can get.
and so on.
Step 3 – Define the Problem
Before carrying out the work connected to this step note down the actual time consumed for carrying the previous step of activities. Note it down in the register for analysis at the end.
Even though people are from the same area it does not mean that all have a clarity about the problem selected. So it is necessary to define the problem. The best way to do is using a ‘Flow Diagram’ showing the entire process related to that area and indicating the problem spot.
Step 4 – Analysis of the Problem
What is the difference between problem definition and problem analysis. Problem definition is explaining where the problem occurs and what are the difficulties and losses incurred in a general way. As we said earlier workmen normally do not have access to the data. Analysis calls for data i.e. number of times the problem occurred, effect of the problem in detail and data on the losses etc. With the help of the Facilitator they should get the data and carry out the analysis. So learning data collection and categorizing them or stratifying them becomes an important activity here.
Step 5 – Identification of causes
The Cause and Effect Diagram helps in this. Since problem is in the work area of the Quality Circle they know the seriousness of it and also have an idea or reason for this problem. The problem is persisting, either because no serious effort was put to find a solution or they were not involved. Many times the solution is eluding them and they are unable to pin point the actual cause.
One important point to be remembered at this stage is problems have multiple causes and variations are due to them. When an analysis is carried out it is necessary to find out all the possible causes. Such an attempt will give a clarity and will help to be cautious when some other cause creeps in at a later date. All causes are expressed verbally. They need to be collected and put in an orderly manner for better understanding and also to arrive at the root cause or causes.
Cause and Effect Diagrams are generally made in three different ways. Since at the beginning people involved may find it difficult to organize them as cause, sub cause, sub-sub cause, sub-sub-sub cause etc., it is better to start the learning process by making Dispersion Method of Cause and Effect Diagram, then going to Enumeration Method and finally to Production Process Classification Type Diagram. Once again our observation is that Quality Circles by and large do not put serious effort to learn how to make a proper Cause and Effect Diagram. They generally prepare one with a lot of inaccuracy, which does not help them to find the root causes.
Step 6 – Finding out the root causes
After making the Cause and Effect Diagram Quality Circle should discuss and arrive at what they consider are the root causes. This may call for collection of data and work spot follow up and study. This has to be decided and responsibility has to be assigned
Step 7 – Data Analysis
Analysis may call for further floor exercises and follow up and making Histogram and/or Scatter Diagram.
Step 8 – Developing solution
Pareto Analysis is a very useful technique, which can be used here. Generally it is very difficult to eliminate a problem altogether. We can only minimize them to a practical limit.
For example a defect level of 6% which exists in a process is considered very high and a reduction to 2% level would be considered appreciable. Once it is brought to 2% one may have to pay attention to another problem that is causing 4% defect. This is a continuous process. Over a period we will once again come to the 2% level problem and efforts will be put to reduce them. (In 1930s Dr. W. A. Shewhart introduced Control Chart using 3-Sigma Control level to improve the efficiency to 99.73% level). It was much appreciated. Now with the intensive competition and availability of advanced control system and excellent equipment, etc. efficiency expected is 99.99966% level (3.4 parts per million as defect) under six-sigma control system.
Generally 80% of the defects are caused by a few causes – may be two, three, four in number. Pareto Analysis helps to identify these vital causes. But it may not be possible to take all these to find solutions. Without understanding the purpose, some of the Quality Circles try to cover all defects coming within the 80:20 ratio at the same time and end up in confusion. They should restrict their efforts to one cause contributing the highest percentage of the problem at first and then go for others later on. However, if one or two causes are interrelated, they may be attempted at the same time.
Once causes are decided the next step is to find the solution. While developing solution to any and every problem, the Quality Circle members should use the technique called counter measure matrix or an action plan which shows time bound and responsibility shared activity in the following format.
The above action plan is found to be very useful in involving every member of the Quality Circle and also for proper follow-up. The QC members may not be able to carry out every action as planned due to many unavoidable reasons. However, the planned date helps them as a reminder to minimise the wastage of time.
Like a problem may have multiple causes, they may have different solutions also. Generally they are arrived at through brainstorming, they need to be shown in proper sequence.
Step 9 – Foreseeing probable resistance
Normally, if we try a drastic change, there would be resistance to that change and people do not easily accept such a change. It does not necessarily mean that there would be a resistance to every change that is brought about. In case a Quality Circle faces a resistance they should mention it and what they did to overcome that.
In case the solution was accepted as it is, they should state that. It is necessary to make it clear that they are aware of such resistance and as a practice they should make a presentation at this stage to their department head as well as to fellow workmen in their area to solicit their viewpoints and suggestions. This would be a positive approach and beneficial to all concerned.
Step 10 – Trial implementation and check performance
It is usual to try the solution arrived at on a trial basis, to a limited area or period and observe the performance of the plant/machinery or system closely. At the same time it is not necessary that every solution needs to be implemented only on a trial basis before final implementation. If there was no need for such trial implementation the Circles may mention it by giving the reasons why it was not carried out.
Step 11 – Regular implementation
Once validity is checked and improvement observed with data regular implementation can be done. One important activity here is incorporating the changes in the working system for validation.
Step 12 – Follow up review
This is an important step since the Circle has to know how the solutions are implemented or are there any further need for changes.
Finally the Milestone Chart showing the planned and actual has to be shown with the relevant data and back up documents as explained.
Mile Stone Chart based on Actual time needed.
At the beginning we have explained how to make a Milestone Chart. The Milestone Chart showing the planned and the actual (in two colours) should be shown here with the reasons for the variations at this place. First data for that:
Source: QCFI HQs publication
Q-Edge 