To day the word “Six Sigma” is being used extensively and this is really related to the quality of any process output. There is always variation when the same activity (process)is repeated. When variation in the performance is in excess, it causes defects and delays. Is it possible to reduce variations and make sure that variations in performance parameters are within the acceptable limits? What should be done to minimize variation? The First step is to identify that there is variation and then understand what caused the variation. Six Sigma methodologies are all about improving the processes to eliminate such variations and perfecting the processes to perform at the highest efficiency level and deliver close to 100% process output quality.
Every process has an acceptable average performance and an acceptable variation in performance. For example, when a customer goes to a bank to withdraw money, he expects that the money will be paid in less than 10 minutes time. If the cash is paid in 10 minutes time, the customer is satisfied and will have a good opinion about the bank. On the contrary, if the customer has to wait for about 30 minutes and still finds that the process is not complete, he will be dissatisfied and might even think of changing the bank.
So ,it is important that the process performance meets the customer expectation and the inherent variation in the process is within the acceptable limits to the customer’s expectation. Every process must be in a position to satisfy the customer and wherever the process does not meet the customer requirement, such processes need to be corrected. It is necessary to know how to understand and improve the process.
As seen in Figure, a normal distribution curve has a positive and negative spread of 4.5 σ. But, in practice,it is observed that every process has a tendency to have inbuilt variance to the extent of 1.5 σ. Due to this phenomena the spread really is 4.5 σ+1.5 σ on both sides of the mean. That means in order to deliver 100% acceptable quality output ,the process output must be within 6 σ (sigma) spread on both sides of the mean as shown
The principle of Six Sigma is near perfection. The same has been explained in terms of percentage of area of acceptance. In case the customer specification is close to +/- 1 sigma then the output would have only 30.23% of acceptable quality and in turn produce 69.77 % of defectives. So under a normal distribution curve the area of acceptability is 68.27% (figure 1-3) and when the correction of spread of +/-1.5 sigma is applied this reduces to 30.23% as shown in figure 1-5.It is very important to understand this phenomenon which calls for 6 sigma capability.
Keeping the above fact in mind, the processes are to be improved to have 6sigma capability, so as to achieve 99.999% of acceptable quality. In other words the number of defects per million opportunities shall be less than 3.4 defects. (DPMO).This principle of Six Sigma quality is applied to all the processes, so as to get near perfect output. Six Sigma quality really means that the process output will meet the customer requirement consistently.
There is considerable confusion in the minds of many people that Six Sigma is a certification process. No. This is a process (methodology) adopted by any organization to perfect its processes to get the highest quality of output and generate maximum customer satisfaction. The goal is to reduce the variation to the minimum and achieve minimum standard deviation (sigma), so that even when the process spread is +/- 6 Sigma to the mean, it will be well within the customer specification.