We recently had a discussion on another platform on this topic. You may like to continue the discussion on Fraternity. Vishwadeep Khatri Principal Consultant, Benchmark Six Sigma "If a process is not yet stable, you cannot apply DMAIC. Six Sigma DMAIC is not suitable for processes which are in a chaotic state" Is this entirely true? Why/ Why Not? Prem Krishna DMAIC is more concerned with improving and controlling the process than developing a new process. We can apply DMAIC for processes which are in chaotic state however DMADV can be a better approach as in DMADV we are more concerned with coming up with a completely new way to produce something than just evolving a system already in place. We can even use DMADV in the place of DMAIC when the process hasn't yielded the results that you wanted in the first place.The main difference between the two is the way that the last two steps are implemented. V.B. Manian Six Sigma consultant at SRPL, Chennai Constant Unstability/Chaos in a process generally refers to a situation which has more to do with innovation, design, etc, where DMAIC may prove ineffective as it has more to do with improving an existing process rather than managing & utilising chaos to produce results. However, DMADV approach under DFSS is best suited to handle this fruitfully. As mentioned by Prem the main difference between the two approaches (DMAIC & DMADV) is the way in which the last two steps under each are deployed. Ramesh Rajagopal TQM Manager at Molex India Limited Good question. The same question can be asked about lean also. There are many reasons for instable process. We can broadly classify them into 4 Ms (Man,Machine, Materia and Method). In my view before using lean or DMAIC , we need to eliminate all the causes of instability to the extent possible using tools like 5S, TPM, Basic 7 QC tools, ISO 9000 etc.. Example : In one of the project people showed how DMAIC helped them to reduce rejections from 30 % to 10000 ppm. Most of the causes in this project were related to stability and does not call for tools like DOE. In my view , bring stability first before going for any improvement V.B. Manian Six Sigma consultant at SRPL, Chennai Well it is obvious & understood that an unstable process needs to be investigated first to reason the instability in it & arrest the same by use of other tools such as Lean, TPM, QC tools, etc, to the maximum extent possible before going ahead with application of SS methodology therein. The point in question here refers to (if I'm not wrong) processes that are inherently dynamic (ever changing) in nature that cannot be controlled & the possibility of application of SS in such a situation. Hope this clarifies the question. Khandwala Saurabh Director and CEO at R B OneSource Pvt. Ltd. On a parallel note, the fact that process is unstable calls for DMAIC which could be a black belt project itself. Inside of that there could be multiple DMAIC/DMADV sub project depending on the complexicity. This would provide structured approach to problem solving Lt Cdr Shalini Agrawal Lean Six Sigma Initiative Lead at TCS Six Sigma DMAIC approach address the common causes present in a process. As we all know that common causes lies in comparatively a stable process, free from special causes which are found more in a unstable state of process. Applying DMAIC for an unstable process will be under utilization of SS tool box to a great extent and will be a waste. VB, your last response has created a confusion rather twisted the question.... I think for a process where unstability is inherent... then that 'unstability' will no more remains as special cause rather becomes common cause for that process but with a higher degree of variation ....'m I making a sense? Khandwala Saurabh Director and CEO at R B OneSource Pvt. Ltd. Yes Shalini, It makes sense, 'Unstable' processes would just throw up higher variations. Over a period of time, 'Unstability' would become common cause for the process. There would be longer 'Measure' phase to capture all the possible variations before it could be improved. V.B. Manian Six Sigma consultant at SRPL, Chennai Shalini, you have tried to raise a valid point wrt to variation cause (being common) in an inherently unstable process. Before, discussing your point further, let's refresh our memory on the meaning of common & special cause variation, in brief: Common cause variation: Common cause variation is defined as one that IS inherent in a process occurring under the SAME circumstances every time in course of the process. So, we can say that the pre-requisites for a Common cause variation are: 1) The variation occurring should be repetitive (& thus inherent) in nature. 2) The variation should arise under the same circumstances every time. Special cause variation: Special cause variation is defined as one that is NOT inherent in a process & occurring under UNUSUAL circumstances in course of the process. So, we can say that the pre-requisites for a special cause variation are: 1) The variation occurring should not be repetitive (& thus not inherent) in nature. 2) The variation should arise under unusual/different circumstances. After clarity on the above, how do we go to define an INHERENTLY UNSTABLE process ? An inherently unstable process is characterized by the below: 1) DISSIMILAR NATURE of activity(ies) (for eg, heating in the laboratory under a DIFFERENT temperature every time) even if the activity(ies) (ie, activity of heating in the laboratory) is repetitive in its/their sequence of occurrence in course of the process. 2) By virtue of the above, unusual/different circumstances are created every time in course of activities (even if repetitive) in the process. From the above, anyone should be able to conclude with reasonable certainty what kind of variation (common cause / special cause) arises in course of an unstable/chaotic process, without the need for me to state the obvious over here. Taking this forward, if we attribute an unstable/chaotic process to innovating / designing something (as purported by me above), one as a SS professional should know whether to use the DMAIC or DMADV methodology. Another golden rule for the success of any SS project is to solve the problem in the soonest time possible (whether using DMAIC / DMADV approach) instead of dragging it over any particular phase (say Measure phase, for eg), unless warranted by concrete reasons for the same per situation in hand, provided the right methodology (DMAIC / DMADV) is chosen to solve the problem. Cheers. Rajeev Kamboj General Manager - Quality Assurance & OPEX Dear VB, Going by your explaination below: "After clarity on the above, how do we go to define an INHERENTLY UNSTABLE process ? An inherently unstable process is characterized by the below: 1) DISSIMILAR NATURE of activity(ies) (for eg, heating in the laboratory under a DIFFERENT temperature every time) even if the activity(ies) (ie, activity of heating in the laboratory) is repetitive in its/their sequence of occurrence in course of the process. 2) By virtue of the above, unusual/different circumstances are created every time in course of activities (even if repetitive) in the process." I don't think we are considering the same process, if the temperature is changing every time, we are changing the process characteristic which means the process can't be termed as same since one or more 'M' is changing out of 5Ms (man, machine, method, material & mother nature) which constitute any process. Therefore I go by Shalini's explanation if the process is having only common cause of variations it will be termed as stable. wbr, Rajeev Kamboj V.B. Manian Six Sigma consultant at SRPL, Chennai Hi Rajeev, Just to reiterate what I meant above (previous to previous post) in other words is, performing EXPERIMENTS (please note that it is plural) to design or innovate something new would be called as the process here. Changing the nature of activities (for eg, heating at different temp every time) in course of the process (ie, for each experiment) in our attempt to design/innovate something new is what makes the process UNSTABLE. So, instability/chaos is the basic characteristic of the process in course of trying to design/innovate something, which (instability) is represented by the changing nature of activities in each experiment under the process. Going by your contention, even if we call every experiment as a new process every time (based on the varying nature of activities, therein), then how do we know whether the cause of variation arising therein is common or special unless we choose to repeat the same nature of activities in that process / experiment over a period of time for observation. So, if we look at things from the view point of a one time experiment/process, we see that EVERY TIME DIFFERENT / NEW / UNUSUAL circumstances are created (by changing nature of activities in each process / experiment) & by that virtue we can attribute the variation arising therein (a one time experiment / process) as special cause variation only unless we choose to repeat the same process/experiment for observation & further conclusion on the cause of variation/s arising therein. Cheers. Srikanth Srinivas A.G.M. - WAREHOUSE at Nagarjuna Oil Corporation limited Beautiful question and interesting responses from the team. I am seeing that many have already assumed certain things; for. e.g. 'stability to controllable', 'constant instability to innovation', 'relating DMAIC process to other QC tools/process/ standardization like 5S, ISO9000' separately, 'removing causes of instability before improvement', 'unstable process throws up higher variations', 'common causes being part of stable process', 'applying DMAIC to unstable process will be under utilization of SS', etc. etc. Well, to add to the interesting debate, pl allow me to suggest few more points to ponder! 1) Pl look closely at Mr. VK's question "If a process is not yet stable, you cannot apply DMAIC. Six Sigma DMAIC is not suitable for processes which are in a chaotic state" Is this entirely true? Why/ Why Not?" 2) 'DMA' is common for DMAIC AND DMADV!! 3) Pl refer Wikipedia for definition of Common and Special Causes 4) Further analyse special vs common causes in the light of 'desirable variations vs not-desirable variations', and ' measurable vs non-measurable causes'. 5) Can one look at stable process as 'process variations with singular central tendency' whereas unstable process as ' process with multiple central tendencies' or 'frequent shift of central tendency'? I hope, the above will lead us to converge on a singular opinion! lakshminarasimhan santhanam Good One!. Amol Dixit Assistant Quality Assurance Manager at WNS Global Services In my opinion, DMAIC can also play a role in stabilizing the process; rather than waiting for the process to stabilise and then apply DMAIC to bring the process within control, the same can be used to stabilize a process. While assessing the process capability, segmenting out the data set based on variation observed and separating out the special causes can assist in arriving at a decision on best use of DMAIC for that process...Thanks. Ramesh Rajagopal TQM Manager at Molex India Limited I just want to add a quote by Walter A. Shewhart before my comments "A phenomenon will be said to be controlled when, through the use of past experience, we can predict, at least within limits, how the phenomenon may be expected to vary in the future" Hope you all agree that Controlled means stable OR predictable OR process with only common causes OR Normal process. Even in medical terminology they use the terms like condition of patient is STABLE (It may be worst but predictable !!). If the patient's condition is not stable they cannot decide on which part to operate !!! Using DMAIC approach to unstable process is like shooting arrows in dark !!!. OR using DMAIC to bring stability is according to me is a wrong approach and we will end with actions like - Control plan corrected, drawing rectified, People trained, maintenance frequency fixed, Instrument calibrated etc.. which are the fundamentals. Finally we may solve the problem but not using the Six sigma tools !! In the process we loose the lessons on importance of stabillity. Team will be rewarded and they become good fire fighters !!! So let us set the foundation right before constructing the building !! I agree with Shalini that "Applying DMAIC for an unstable process will be under utllisation of SS tool box to a great extent and will be a waste" Ramesh..... V.B. Manian Six Sigma consultant at SRPL, Chennai @Srikanth The reason for process instability needs to be looked into. If it can be arrested then same should be done before applying SS DMAIC, as opined by some of the above respondents. If instability is an inherent characteristic of a process, which involves design, innovation, etc, then SS DMADV holds more relevance. This is largely relevant per our understanding. We have variation both in stable as well as unstable process that can be further categorized into desirable & undesirable. If you know or have come across some meaningful information on the topic under discussion as purported by the points raised in your above post, would appreciate you sharing the same here for the benefit of one & all, as I believe we are trying to have a constructive discussion here with valuable inputs from those who choose to opine more than playing a game of treasure hunt. Trust you won't take me wrong & look forward to your insightful response to enlighten us further. Cheers. 16. Sujeet Singh Project Leader-Transition (Program-Implementation) at Etech, Inc. I would like to share one example from one of my project. it was on call drops in an outbound process. Call drops was happening because of following reasons. 1. Number of CSR available to answer the call. 2. Dialer hit rate 3. Dialer Ratio per agent 4. Contact rate The first four causes are controllable but last cause is uncontrollable because it depends upon lot many factors from customer's availability stand point with respect to day/time/occasion if any; Which is uncontrollable. In this situation because of undesirable contact rate fluctuation we were getting higher call drops. Now to keep call drop at a desired level we implemented real time monitoring of contact rate and accordingly make the changes in the dialer settings with respect to hit rate & ratio. This process was in chaotic state however after following DMAIC steps, now we are getting the desired result. So in my opinion; we can implement DMAIC for the process which are in chaotic state but again it is situational call to make... wbr Sujeet