There are several scholars and writers who have explained Six Sigma in different ways. Some of them explained that six is a problem-solving methodology, where it talks about an equation which is Y=f(x). Here in this equation Y is for the metric which needs to fixed, x is for the potential cause, due to what the outcome/output in the form of “Y” is coming out and f is here for the function of x. So by this equation, statisticians tried to explain that instead of trying to fix the problem in form of “Y” superficially or temporarily we try to identify the potential causes due to what our metric of “Y” is not performing well. Then we collect the data around all collected potential causes and validate the statistical relationship between all “Xs” and “Y” and identify the vital “Xs”, then we prepare action plans on these “Xs” in such way that my “Y” gets permanently improved.
Some of the statisticians have explained six sigma is as a philosophy, the meaning of philosophy is conception, where it talks about some basic need to complete a project successfully, these basic needs are customer focus, engagement with bottom-line employees and to reduce variation and removing defects, further it explains that if you want to fix your “Y” which is a metric with issue/problem, to remove this problem you have to engage with the bottom line team as only they are the right point of contact to know the correct causes about the bottlenecks in the process. Being a customer facing they can give you a correct reason for why the metric “Y” is not performing well.
Further, it was explained that Six Sigma is a metric, being a metric it explained that with a common yardstick or unit we can measure the capability of any process irrespective of the type, size, and nature of any process. As that common yardstick or unit is Z score/DPMO. As per Six Sigma the thing you can’t measure you can’t improve. It talks about that Six Sigma is a data-driven methodology; if you do not have a data you can’t improve any process. Six Sigma methods can only be implemented for a metric “Y” with data and further you must collect data around all collected potential causes.
Six Sigma is also explained as a management system, where it talks about 3 popular frameworks which are DMAIC, DMADV and DFSS. The most popular framework is DMAIC which is used to fix some issue around any metric in the existing process, DMADV is a framework which used to while launching a new product or process and DFSS is used for automation purpose to any process. DFSS is a framework which can be used either with DMAIC or DMADV framework; it is only used when an existing or new process has achieved 4.5 sigma levels. Another prerequisite is to automate any “Lean Implementation”. Lean must also be implemented before automating any process. Any existing process can only be automated during “Improve Phase” of DMAIC framework and if it’s a new process then it can be automated during its “Design Phase” while working on DMADV framework.
Six Sigma is a word which talks about 6 standard deviations. Any process is known whether it is running on six sigma level or not if 6 times standard deviations are getting fit between mean and considered specifications limits. If any process contains both side specification limits which are known as USL(Upper specifications Limits) and LSL (Lower Specifications Limits) then 6 standard deviations must be fallen between both sides of mean value towards USL as well as LSL. If any process contains only one-sided specification limit either it is USL/LSL then 6 standard deviations must be fallen between the mean value and considered specification limit to that process. As per defined Z score table, if any process is running at 4.5 sigma levels, the number of defects per opportunities will be 3.4/1000000, where 3.4 are the number of defects and 1000000 are the number of opportunities. But this 4.5 sigma level is known as a Long-term Z score, whereas we always measure the capability of any process in Z scores short term. To calculate Z score for short term we have to add 1.5 to Z long term. Sigma level of any process/industry is decided to be run as per the criticality or severity on its end user. Means if the product will fail how severe or the critical impact it will leave on its end user/customer. So this is not mandatory for all the industries to run their processes on 6 sigma level.