# 7QC tool and Its Uses

## Problem solving by 7 Quality Tools

Quality pays for itself. Without good quality no industry can survive competition. So it is very important to sustain quality of products. In this article 7QC tools explained which are widely used by automotive industries to short out problem and closing that by taking appropriate countermeasures.

### What are Seven Quality tools?

Basically 7 quality tools are problem solving tools, which are widely used in automobile industries. These problem solving tools not only used in Automotive industries but also in non-automobile sectors like service & Retail sectors.

These Quality tools work on PDCA cycle. PDCA stand for plan-do-check-act. For solution of any quality problem we follow funnel based approach, in which we identify problem severity and then define that problem (explain the real problem), we find possible causes (by using fishbone diagram) and then we find root cause of problem, then we take countermeasures then after we evaluate that countermeasure (Here we check effectiveness of counter measures) and finally we standardized it in documents for horizontally deployment for other parts.

### List of 7QC tools

1. Check Sheets

2. Pareto Chart/Diagram

3. Stratification, chart & PFD

4. Fish Bone Diagram/Ishikawa Diagram/Cause and Effect Diagram

5. Histogram

6. Scatter Diagram

7. Control Chart

### Why need of 7QC tools?

Because companies with Quality program (7QC tools, Quality Circle, PDCA, QMS, Kiazen activities) makes more profit. The 7 QC tools provide common methods of analysis

to help problem solving teams operate effectively. It helps in taking decisions faster and objectively. Return on investment is increases. By implementing Quality tools following benefits can be obtained:-

- Improve Quality
- Decrease cost
- Improve Productivity
- Decrease/ Increase price
- Stay in business
- Provide more jobs
- Retune on investment

**Basically there are 7 steps of problem solving:-**

**1. Define the problem**

- Explain the problem [Details of problem]
- Project the losses [Effect of problem]
- Identify who will work on this [team to solve the problem]

**2. Data Collection**

- Collect the necessary data by using check sheets

**3. Analysis**

- Use appropriate QC tool for analysis like- control chart, pareto chart, fish-bone diagram, Histogram, scatter diagram, Flow diagram, etc.

**4. Action**

- Action is Counter-measure find out after analysis

**5. Study**

- Study the impact of action taken/Check effectiveness

**6. Standardize the solution**

- Standardize mean, we have to record it in documents for past learning, same countermeasure to be taken if new development, or Horizontal deployment in other similar process at different location/Country

**7. Continuum**

- Continue to look for other problem of shop floor and provide solution for the same.

### Check sheet in 7QC tools

Check sheets are paper forms in which we collect data by writing or by printing. In check sheet we fill both attribute and variable data. Examples of check sheet could be Defect sheet, Process check sheet, production report, route cards, Kanban card, Record of bar code/Packing slip on parts could be check sheet. Sometime there could be separate check sheet for each operation for recording purpose, which can be used for analysis.

Purpose of Check sheet:- Collecting data and using it as input to determine source of problem. By check sheet data analysis is done, we track basic or advance level of source of problem by checking-who worked, what was date, what was parameters, who give approval, what was time and production as per check sheet. So these information give information about source of problem.

### Pareto Chart in 7QC Tool

**Definition:-**

It is combination of column and line graph, differentiating major factors contributing to the problem from the other factors which have less contribution. Or we can say it is technique to segregate vital few from trivial many.

Pareto chart or pareto diagram is used to prioritize the problems, by this we try to focus on key problem. Preto chart isolates (Separate) vital few from trivial many. It is also 80:20 principle, because 80% improvement can be achieve by working on 20% of causes.

Preto chart/Diagram was developed by Vilfredo Preto of Italy.

#### How to make Pareto Chart?

Before making pareto chart we must have a data sheet, which contains list of problem classifications. Following steps to make pareto chart:-

- Arrange each problem classification in order from highest to lower in excel.
- Write down %age of overall total in next column.
- Write Cumulative %age in next column [for 1st defect defect percentage will be same as %age of overall total, for 2nd it will be sum of %age of overall total + cumulative %age of 1st one, and so on. Excel formula could be
**=$E2+D3**, and the scroll down. - Select data sheet and insert column chart.
- For Cumulative %age-change chart type to “Line with marker”.
- You can delete %age of overall total by clicking in excel

Here is Check sheet in below picture:-

Preto chart based on above data sheet will be like this:-

Left vertical axis of pareto chart has count or cost as per data. Each vertical bar represent contribution to the total, the bar at the left has highest contribution to count and cost. Each right vertical axis represent percentage demarcation. A cumulative line is used to add %age from each bar.

From above pareto chart photo we can have two understanding, first- Identification of bar which is contributing the most, here it is Dent, second- by cumulative line we determine how much of total problem will fixed by addressing the highest few.

Here pareto principle is 80% results are determined by 20% of the causes, therefore we should try to find 20% of defect which are causing 80% of all defects.

##### Conclusion of pareto chart

So if we take care of only Dent defects we can take care [or Reduce] of 50% problems, and if we take care of both Dent and scratch we can take care of 70% problems.

### Stratification, Flow chart in 7QC tool

**Stratification** is process of data arrangement, in which we separate the data so that pattern can be seen. Because when we collect data from different section and analysis on the whole, then we can not get purpose of analysis. So we use stratification for data arrangement.

Here we can collect data and can be stratify in following cases:-

Materials, Suppliers, Equipment, rejection types, date, months, Locations, Parts Types, Departments, operators, etc… Ultimate purpose is we not only have to separate data but also have to diversify with proper breakdown.

Flow Chart is visual illustration of sequence of operation, process to make a product. In flow chart we not only give name or number for process we also give some symbols which indicates action being done for different operations.

Graphs are pictorial descriptions of any data. Some of Bar graph are Bar graph, Line graph, Pie graph, Radar Graph. By pictorial description we can summaries data in very short also long data can be explained very easily by pictorial description.

### Fishbone Diagram/ Cause and Effect diagram/Ishikawa Diagram

Kaoru Ishikawa san made fishbone diagram. Simply known as Cause and effect diagram, which shows relation between problem and its possible causes. In this diagram we arrange all possible causes which could raise a problem.

Purpose of fishbone diagram is to find out root cause of a problem. Why why analysis is also part of fish bone study.

For fishbone diagram we arrange 6M (Man, machine, method, material, measurement and mother-nature), then these 6 source of variation is divided into sub-group and so-on.

#### How we use fishbone diagram to find root cause of problem?

Step-1:- Make fishbone diagram and write down all possible cause of problem in 6M

Step-2:- Find suspected cause

Step-3:- Find Valid Cause by validation of suspected cause (Simulation to be done in validation)

Step-4:- Why why analysis over valid causes for root cause

When we do Why Why analysis we try to root cause at both stage, occurrence and outflow. Why defective part generated and if generate why not detected?

Countermeasure will be taken on both.

Mostly 5th or 4th why could be root cause, but it differ case to case.

### Histogram in 7QC tool

Histogram is graphical representation in bar form in which we check frequency distribution of one continuous variable within collected data.

By histogram we can check following:-

- Future performance of process,
- Identify change in process
- Predict whether production lot with this process can meet customer requirement.

**Histogram is 7QC tool by which we can see process shape, process centering and process speared. **

#### How to make Histogram?

Before making measure or collect data of 50 to 100 nos. So lets understand it with example, We have tube of width 200 mm tolerance is +/-10 mm. So max value will be 210 mm and Min value will be 910 mm.

Observation were taken for 100 times and observation recorded which is as below:-

- No of Bin/No of Bar/No of Classes [All same thing]
- Width of Bin/Bar/Class
- Class Boundary

- No of Bin/No of bar/No of Classes are dependent upon number of data measured. It could be represented by [k], generally it is square root of Number of data. Here number of data is 100, so No of class/No of bin/No of Bar would be 10.
- Class width can be calculated by Dividing Range [Max observation-Min observation] by No of Bins. It will be 220-190=30, and 30/10=3.
- Class Boundary is table in which first value is minimum observation, which is 190. Second will be 190+3=193, and go on adding till all measurement placed in their classes.

For frequency count we put formula of frequency, **=FREQUENCY(A3:J12,D17:D27), **here D17 is 190, and D27 is 220.

Then to calculate rest values below first cell. Follow this trick, Select all cell and click in formula bar, press ctlr+shift+enter. All values will be automatically come in cells.

From this Histogram it is clear frequency distribution is not normal, process is not OK. Variation are huge, which will cause failure.

Now see the result of second trial, observation are below:-

No value is out of spec, maximum was 210, and minimum was 190. Here we calculate frequency count for histogram, we obtain below result:-

And histogram will made on the basis of above data, for that we will first insert bar chart, then we will select data and scroll down under frequency count. Next will be class boundary on x-axis, that is all.

From the above Histogram, we can see frequency distribution is normal, which indicate variation is zero, and values are more about mean.

Now i have repeat normal distribution 2 times here, so what is this?

To understand it, we first have to understand standard deviation.

Standard deviation is amount of variation within data range. More standard deviation mean more variation and less value of standard deviation is less variation. It is represented by sigma. It is generally spread, which indicate average distance of individual number with respect to mean. if we have to calculate standard deviation of 1, 2, 3 mean will be 6/2=3 and standard deviation will be under square root (1-2)²+(2-2)²+(3-2)²/3 = 1.

Standard deviation is square root of variance, which is sqr root of (X-mean)^2/N.

In above data mean was 201.2, variance was 22.84 and standard deviation is 4.80. Here upper control limit is 210.and lower control limit is 190. We see that in 2nd trial frequency distribution is within 2 standard deviation. No value is going within 3 standard deviation. Which indicate process more about the mean.

### Scatter Diagram or Scatter Plot in 7QC tool

It is a tool to identify possible relation between two set of variables. One variable will be dependent and other will be independent variable. One variable is plotted on the horizontal axis and other on vertical axis. The pattern of their intersecting point can graphically show relationship pattern.

If the variable are correlated , the points will fall along a line or curve, stronger the correlation, closer the points will be to the line.

This tool is used after making fishbone diagram, to identify relation between cause and effect. Scatter diagram was invented by sir Francis Galton.

Here is data recorded of two variable one is HRB [Hardness] and other is TS [Tensile strength].

Scatter diagram will be plotted between HRB &TS, HRB will be on X-Axis and TS will be on Y-Axis. Here just ignore Red values in above picture.

From the above scatter diagram we see there is positive correlation between HRB and TS, since all points are closer to line. Various types of scatter plot is as below:-

- If R>0-it indicate there is positive linear relation
- If R<0- It indicate there is negative linear correlation.

Generally R varies between 1 or -1. R can be calculated by formula. Which is SS(XY)/Square root [SS(xx). SS(yy)]. SS is sum of squares.

If we calculate R [coefficient of correlation] by using above formulas and data values. It come out 0.75 which is more than zero, which indicate correlation is positive linear.

But there is not always correlation between two variables, sometime you can judge easily by Marely see the comparison, here is an example- Level of cleanliness and number of infection. It is negative relation.

### Control Chart in 7QC tool

This is tool by which we can detect abnormal trend with the help of line graphs. Control chart set limit of variation, we have to control process or variation within the control range, if anything going out of range, that is abnormal.

Concept of control chart was made by Dr. Walter Andrew Shewhart in 1924. In every control chart there will be center line [Average value], Upper control limit=Average+3sigma and lower control limit= Average-3sigma.

- If no value go out of UCL & LCL.
- Most points are close to average
- Equal number of point below or above center line

**If you like my work, please support by sharing and taking subscription on social platform, If you have suggestion and advance please let us know. **