A Scatter diagram or scatter plot helps to find the correlation of two variables. One of these is a dependent variable which is plotted along the Y-axis while the other independent variable which is plotted along the X-axis.
Once the correlation is determined you can predict the behaviour of dependent variable based on the independent variable. The control parameter is independent variable which influences the dependent variable.
Example: We can study the correlation of the number of hours of study by a student per day to the marks scored in the examination.
Types of Scatter diagrams
Scatter diagram can be classified as under based on the correlation.
- No correlation
- Moderate correlation
- Strong correlation
The data is spread randomly with no relationship between the two variables.
The data shows that the two variables are moderately correlated & there exists a relationship between them either positive or negative.
The data points are so closely related to each other that they form trends like lines or curves showing that they are strongly correlated
- Relationship of two variables can be determined
- The range of data flow can be studied
- Non-linear patterns can be easily visualised
- Very easy to plot
- Exact extent of correlation cannot be found
- The relationship of two or more variables cannot be studied together
You can find the relationship between a cause and effect or between two causes. The type of relationship can be identified. To analyse the trend of relationship one parameter can be changed & monitor the other.
COST OF POOR QUALITY (COPQ)
Cost of Poor Quality refers to all the costs associated with the problem.
There are many Cost of Poor Quality (COPQs) that are easily recognized such as rework, inspection, testing, customer returns, rejects and complaints etc. However behind these there are many COPQs those have become so common that we have started viewing them as normal, such as Inventory, late payments, expediting costs, high employee turnover etc.
There are four categories of COPQ:
This refers to any systems, processes or procedures that exist only to look for problems, such as inspection. Examples like, Verification, Audits, Ratings etc.
There are also so many systems, processes and procedures to prevent things going wrong. However they are beneficial, they are still costs of poor quality. Examples like, Quality planning, Quality assurance, Training, Incoming inspections etc.
Problems that occur within an organization and they do not affect the customer directly but still they should be seen as Cost of Poor Quality. Eventually these COPQ’s will reach customer in the form of higher prices or delays. Examples like, Waste, Scrap, Rework, Failure analysis etc.
An external COPQ id the cost of any defect that reaches customer, they can become very significant. Examples like, Repairs and servicing, Warranty claims, Customer complaints, customer returns etc.
Internal failure costs and External failures costs known as cost of poor quality and Appraisal costs and Prevention costs known as cost of good quality.
A Histogram is a frequency distribution curve of continuous data. This graph helps us in interpreting the data underlying the distribution.
Histogram example is given below.
Marks scored by students in a class in mathematics subject.
How to construct the a Histogram
- Collect the raw data for the study
- Categorize the data in to intervals
- The frequencies are plotted along the Y axis & marks scored on the X-axis.
- Each category or bin accommodates the number of frequencies of marks scored by the students with in the range.
In the above example the marks obtained by the students have been split in to categories of 10 marks each.
There will be no gaps between the frequency bins. The reason being that the Histogram represents continuous data set unlike in a bar chart where you can find gaps. In Histogram the area of each bar represents the frequency and not the height.
The main difference of a histogram from bar chart is that the histogram is that the histogram plots the frequency of occurrences in a continuous data where as a bar chart depicts the frequency of ordinal and nominal data.
Advantages of Histogram
- Performance output of a process can be studied
- Analyse the data to meet the customer requirements
- Output of two or more processes can be compared
- Data distribution can be effectively communicated to all the stakeholders
After constructing the Histogram the distribution of data is analysed based on the shape of the distribution like normal, skewed & double peaked …etc for further course of action.
CAUSE AND EFFECT DIAGRAM
It is a graphical tool that helps to list all the possible causes of a problem/effect.
Cause and effect diagram analysis is introduced by Kaoru Ishikawa in the 1690’s.This was then published in his book in 1990. This Cause & Effect diagram is also known as Ishikawa Diagram / Fish Bone Diagram.
Advantages of Cause & Effect Diagram.
- It helps to determine the root causes
- Encourages team work
- Systematic approach for problem solving
- All the possible causes are understood
- Enhances the process knowledge
Developing a Cause & Effect Diagram.
Step 1- Clearly define the effect/problem. Write down the effect on to the right hand side and draw a box around it. Extend an arrow from the left hand side towards the box at the centre to form a spine.
Step 2- Identify the main causes that are contributing to the problem. All the possible causes can be categorised in general as below. However we can categorize as per the actuals.
Step 3 – For each of the major branches all the possible causes are identified by brainstorming. Possible sub causes for each of the causes if any are listed with more details. This can be achieved by using five why tool.
Step 4 – Once all the possible causes are listed down in detail, now it’s time to analyse the root cause for the effect.
The Pareto chart can be used for further analysis to identify the root cause & find the appropriate solution for the improvement.
Lean Kaizen is a term for a continuous improvement philosophy in Lean. The common name is referred to as a Kaizen and is a Japanese word which literally means Change (Kai) for the better (Zen). It is based on a philosophy of continuous improvement within processes in Businesses across all sectors and industries and involves all employees from the CEO to the Operators on the shop floor.
The term Lean Kaizen or Kaizen, is generally referred to when dedicated groups get together to make a focused improvement together. In reality, although small improvement projects are normally referred to as Kaizens, the term denotes any small incremental improvement in the business.
In the true sense of a Lean Culture, all employees are tasked with making improvements across the business. In fact it is their job to find improvement ideas in search of Kaizen – Change for the better.
The Lean Kaizen method consists of the following principles:
- Personal discipline
- Improved morale
- Quality circles
- Suggestions for improvement
The interesting thing about Kaizen is that It is continuous. It is ingrained in the culture of the Japanese and in some established companies like Toyota and Canon, for example, there are around 60 suggestions per employee every year, which are implemented, shared and resolved. This stance is based on the conception that It is better to make little changes on a regular basis and therefore, “Ten small improvements are better than one large improvement.”
Suggestions are not limited to a specific area such as production or marketing. Kaizen is based on making changes anywhere that improvements can be made. Western philosophy may be summarized as, “if it ain’t broke, don’t fix it.” The Kaizen philosophy is to “do it better, make it better, and improve it even if it isn’t broken, because if we don’t, we can’t compete with those who do.”
A Pareto Chart is a frequency distribution graph where in the defects/errors are arranged in a descending order from left to right to prioritize the most significant causes & to understand the cumulative effect The height of each bar represents the frequency.
Pareto Chart was formulated in 1987 by an Italian economist Vilfredo Pareto which became popular by the name 80-20 rule & from his analysis that 80% of the wealth lies with 20% of the people. Being universal this relationship is applicable to any field.
How to Construct a Pareto Chart
First collect all the relevant data & categorize them into groups.
Label the left side of the Y-axis as Frequency.
Label the right side of the Y- axis as Cumulative percentage
Categories of data are arranged in an ascending order of the factors starting from the left side on the X- axis. Then a cumulative frequency line is plotted which is ascending and flattens off at the end.
The height of each bar in the chart represents frequency of defects. From the chart one can visualize the critical factors which vital few are differentiating from the rest which are trivial many. Hence it is called vital few & trivial few technique. The cumulative percentage line is used to estimate the contribution of each level of factors/defects.
The focus is then shifted to vital few factors & further analysis is carried out to find the best solution and then implemented making significant difference in the process improvement.
It is a four step iterative method for any process improvement. PDCA is a closed loop process where in frequent changes are done on regular basis for continual improvement.
The PDCA concept was initially designed by Mr. Walter Andrew Shewhart in 1930’s but it was popularized by Mr. Deming.
The acronym for PDCA is
1) Plan: At this stage a problem/opportunity is identified for improvement. Start with defining the problem completely & set goals to be achieved. Five why’s method can be used for identifying the root cause of the problem.
Collect all the relevant information, generate & select the best idea and develop a plan for implementation.
2) Do: After finalizing the best solution test the process with a pilot batch. This will help to implement the proposed changes with minimum risk. Adopt a practical approach for implementing the solution rather than perfection & make a note of the results obtained.
3) Check: During this stage the results obtained from the implemented action plan are analysed and compared to the initial expectations as planned. Try out as many trails as possible by repeating Do & Check till satisfactory results are not obtained.
If the results meets the expectation, then proceed to the next step if not repeat the cycle.
4) Act: Here the action plan is implemented at a full scale and monitored continuously for any variation. Since PDCA is a closed loop the process with the new changes is set as a new standard and has to look out for further improvement continuously.
Total Quality Management ( TQM ) is defined as a management approach to achieve customer-satisfaction through continuous improvement of business operations. It focuses on all employees’ involvement to work towards the common goals of improving product quality or service quality with fact-based decision making.
The main elements of TQM are;
1) Commitment by the management: Commitment from the management is the key factor & by adopting strategic and systematic approach to achieving an organization’s vision, mission, and goals the long-term customer satisfaction can be achieved.
2) Customer-focused: The customer ultimately decides the level of quality of the product/Service. Customer driven standards, never compromise on quality are some of the key points in achieving the customer satisfaction.
3) Employee Empowerment: All employees’ involvement in achieving the common goals of the organization. Employee commitment can be achieved by training, recognition and suggestion scheme.
4) Continuous improvement: Continual improvement drives an organization in finding ways to become more competitive and more effective at meeting customer expectations.
5) Process-centered: Process approach is the main aspect of TQM. This process approach drives preventive actions with a consistent output. The process is then monitored continuously for its performance in order to detect any unexpected variation.
Benefits of Total Quality Management TQM
- TQM drives continuous improvement culture.
- TQM helps in achieving long term goals through customer driven approach.
- This will increase the awareness of quality culture within the organization.
- Teamwork will be promoted.
This is a simple and powerful tool to find the root cause of the problem. Sakichi Toyoda the Japanese industrialist has developed the technique in 1930’s.
The five why technique is most effective when the team from the same problem area are involved in finding the root cause of the problem.
The method simply means to drill down to find the root cause of the problem by asking why five times. Once you find the root cause, preventive actions are taken to prevent the problem from recurring.
The five why analysis is most effective for solving simple & moderately complex problems.
Following are the steps to be followed in using five why’s:
1) Form a team who are directly involved in the problem. Appoint one person to be a facilitator.
2) Define the problem precisely.
Ex: Spilled oil on the shop floor
3) Start with asking first “why”
Ask the team why the problem is occurring to find the actual answer without any guesses. Note down the answers.
4) Ask why again four times
Drill down to find the root cause by asking why four more times for each of the answers noted above.
5) Now when we have found the root cause of the problem which is Supplier selection criteria you can prevent this by modifying the procedure for supplier selection criteria. After this keep a close watch on effectiveness of the countermeasure implemented.
It is also called as Poka-Yoke (a Japanese term) that means “Mistake Proofing”. A Poka-Yoke is any mechanism in a process that helps an operator to avoid (Yoke) mistakes (Poka).
It comes under IMPROVE phase of DMAIC.
There are two ways to apply Mistake Proofing effects:-
- Prevention – A Mistake-Proofing effect that avoids mistakes before they occur.
- Detection – A mistake-proofing effect that identifies mistakes once they occur.
Mistake Proofing might also be thought of as an extension of “Design FMEA”.
While an FMEA helps in the prediction & prevention of problems, Mistake-Proofing emphasizes the Detection & Correction of mistakes before they become defects. Defects that may subsequently be delivered to the end customer.
Example - when to Use Mistake Proofing:-
1. When a process step has been identified where human error can cause mistakes or defects to occur, especially in processes that rely on the operator’s attention, skill or experience.
2. In a service process, where the operator can make an error which affects the output.
3. Machines & Processes fail & make errors.
Steps to Implement Mistake-Proofing
- Obtain or create a flowchart of the process. Review each step, & find the concern step & all the possible chance of occurring error.
- Analyze all these user scenarios by asking 5-Why’s questions to understand the ways these scenarios failed or can fail.
- Once you identify the ways these scenarios can be wrong, design & apply Mistake-Proofing technique to avoid all the problems.
- Make sure the technique designed to avoid the defects is working properly by giving errors or warning message for incorrect input.
- Check if this technique has really prevented or caught defects when happening.
Examples of Mistake-Proofing:
SIM card slot in cell phones is designed in such a way that user is allowed to insert SIM card in correct way only. There is no chance for user to make a mistake while putting SIM card in a cell phone. This makes the design mistake proof.