Jump to content

Anil K

  • Content Count

  • Joined

  • Last visited

Community Reputation

0 Average

About Anil K

  • Rank
    Advanced Member

Profile Information

  • Name
    Anil Kumar
  • Company
    Benchmark Six Sigma
  • Designation

Recent Profile Visitors

1,373 profile views
  1. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Danish Khan You join a company where the management seem keen on setting up processes. You start with a bang and the senior management seem to be in line with you. Over a period of time, you start documenting process, drawing flow charts and building on your SOPs. However, you suddenly realize that the middle management has found a way to create deviations in the process by taking special permissions from senior management. There side of the story, "If there are no deviations, we will not be able to do it". The management starts buying it, only to see immediate results and to keep the E-Sats high. Gradually you see that people have stopped following processes and that all your documents are merely an eye wash. What should be done in this situation? How can we pull people towards following processes, after all talks, gyaan and rationales have failed? Raman Bansal I have a different opinion about this culture. I will try to add two cents of wisdom here, however you may still differ in opinion. First, deviation is a not an evil. If the process is working better with the deviations, which means it's very much required to be part of the system. Second, coming to the very purpose of putting processes in place is to establish a standard way of working. Now, stitching above two statements will tell us that deviation is just a scenario which we generally miss to document in the first instance. Which means any process when being followed in normal circumstances shall follow the standard process, however if there is any urgency/fallout, it shall get approval of respective hierarchy in the standard deviation approval system (which again becomes a deviation process). And this deviation count shall get reviewed in monthly management review to discuss how many such deviations were passed and what were the root causes. Let people start working on those root causes and eliminate the need of such deviations in future. A structured system well created can make culture changes gradually, if not immediately. Geetika Moudgil Sharma totally buy above point by Raman. Mohit Sethi In agreement with Raman Bansal point Parthiv Chatterjee Agreed to above reply from Raman..adding to this I would say that deviations are necessary to show us where we can improve and make the SOP more robust.. This is kaizen... Also note that changing trends in Operations or market, always calls for deviations and revision of processes... Navnit Goel Center point to be kept in mind that processes are created to serve the project and not the other way. Agreed upon deviations are necessary for running the project smoothly. However, there needs to be structured and defined way of getting deviations and also a proper review. Processes need to be adhered in letter and spirit. Management should take serious if the processes are not being followed up. First of all management itself should be convinced that processes bring higher productivity, compliance down the hierarchy can be brought with proper guidance, training and counseling. Akshay Kapoor One can have better controls in place with constant monitoring to ensure processes are followed and not just documented but having said that constant review of the flow needs to be checked as changing old methodology is equally important to put a step in future with better results. And as very well said above each deviation to the standard should be measured with proper root cause analysis to eradicate the issue. Maheshwari Subramania This is becoming quite common with many companies, saying we are going agile, but in the name of agile, don't have a proper tracking mechanism (read processes) in place which takes a toll on employees. At the end, we end up in countless rework effort, which is again a vicious cycle. Prateek Kumar Relax, it's a universal phenomenon. Processes are not cast in stone. Before we look outwards lets look inwards.If for a particular process lot of deviations are happening, then probably we have defined the process narrowly in the first place. Secondly, deviations once taken should be taken as inputs to evolve the process or gain consensus to handle such deviations in future.Over a period of time the deviations will go down.Deviations are only an indication that the process maturity process need to be taken to the next level. Shrinivas Gardas Generally, there is a global practice of obtaining approvals for the deviations from the process. Lets have a look at from a output of performing an activity which is deviated from process. Deviations are acceptable till the time it doesn't cross the defined tolerance limit moreover should not affect key perspectives viz. cost, customer, schedule, legitimacy and other financial assertions viz. accuracy, completeness, segregation of duties etc., Having said that, there should be a regular conduct of concurrent monitoring of operating condition deviations and process dyanims which otherwise one will never have any control of the processes with innumerable deviations.... Indresh Saluja What are trying to achieve by process documentation? Are we saying process documentation will supersede the thinking of an individual doing the process? Or is documentation such a robust way that it can cover all deviations and exceptions? I agree to points raised by Raman and Prateek. I have done process re-engineering and documentation with lean across several large companies (service, asset& service, pure asset, manufacturing& pure trading companies). There are not more than 5-6 core business processes( order to cash, financial planning & control, raw to finished, customer request to fulfillment, concept to produce, hiring to exit, customer life cycle management, product life cycle management etc... We present these in one sheet per processes cross-functionally without decision boxes. It's a one shot snapshot which everyone wants as its a good reference point. Highlight boxes where error occurs not where error is seen. Create automation and a metric. Create guidelines for common exceptions as annexure so process is not disturbed. It becomes easy dynamic doc. Check more details in article documented by my team at isixsigma.com. http://www.isixsigma.com/implementation/basics/process-documentation-a-modern-approach/ Shrinivas Gardas Thanks Indresh for sharing ur valuable inputs: a traditional and new way of documentation approach is very well distinguished and explained ... Gr8 work by your team Indresh Saluja Thanks Shrinivas. It has helped us immensely in creating quick changes across several organisations !! Femi Obiomah Before process documentation, please get everyone involved in the process on the same page with you. When developing documents like SOPs, it is helpful for others to know that they were involved in its creation even if they have no inputs. Some of their inputs may sound stupid, but when there's a cosensus on why the idea should be rejected, it makes life easier. It could turn to stormy sessions, but from experience, the outcomes have been fantastic.
  2. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Vishwadeep Khatri Volkswagen did illegal cheating. There are others who are doing it legally. There is 50% difference in emission levels on road as compared to test conditions for some manufacturers. What are your views on legal and illegal ways of cheating? (The driving experience and fuel efficiency get degraded when emissions are kept within limits. Car makers shall probably continue to play on emission norms as compliance goes against buying criteria) Steve Borris Consultant and Author According to the TV news, the tests are followed as the rules define. It would seem the errors are with the rules. I read somewhere that parts, like seats, were removed for fuel mileage tests in some cases (to reduce the weight). If true, that is probably a legal misrepresentation, too. If no one objects, nothing will ever get done. The power lies with the customers. Steve Dr.Narasimhan S., M.S, Ph.D Yes Mr.Khatri....a cheating is a cheating...can we classify this into legal or illegal?..like good terrorism and bad terrorism..?..and I fully agree with Steve that power lies with customers...so only these corporations come up with so many intelligent ways of cheating.... Thomas Prasad Good question Vishwadeep ! All those clinically proven, doctors recommend, so and so certified, best in market, analysts suggested are different side of the same story. Why naturally we get oriented to these words; there is something about the product or service that needs a different story telling for it to be sold. Have you ever wondered .....exclude last 50 to75 yrs we did not find these statements for buying gold, silver or any perishable good like food and other necessities. In today's world all these belong to the category of making money by what ever it takes. Manu Kumar Yes, We all should follow Norms, which is meant for goodness of Society and environment. Norms are not made of individual benefits. Cheating really hurts when it is harming the humanity or nature. Yes, we humans always strive for development but the outcome contains both good and bad things,, we should accept and follow good and minimize the bad things which may arm the society. Ravi Vaidiswaran Company has created the trust of the customers. There is no question of legal or illegal. Kiran Varri Cheating in either form is Unacceptable !! Let's keep #HumanGood above all !! Himanshu Rathore business needs to be sustainable above all. ..d emission levels can be measured and a carbon tax can be levied on manufacturers fr exceeding the limit.for eg :car A has higher emissions than that of B...so A should be made dearer than B in the market. ..That would rein in the manufacturers.Trust or CSR can't hold on for too long unless it is directly linked to the profits..which is why we do business, right! Vishwadeep Khatri Hi Himanshu, I do not think Volkswagen was fighting a sustainability battle. It was a blatant cheating effort to gain global market leadership. Rohit Sharma I don't think there is anything called "legal cheating". This is where ethics of the board and corporate social responsibility comes into play. Naresh Sharma They are not the first one for whom Greed overtook their desire to sustain a Business. Only future will tell us how they will come out of their shoddy practices... Vishwadeep Khatri Thanks all for sharing views. Legal Cheating - Making sure that no laws are broken while the intent of the legal requirement is flouted for business advantage. Example - Mileage reported as higher - This is being done by many vehicle manufacturers using subtle methods that ensure that performance is far better in test conditions (which gets reported) as compared to on-road performance (which never gets reported). Illegal Cheating - Circumventing a legal requirement with conscious intent and using methods that can be proved to be against law. Example - Software usage to switch off emission control mechanism during testing. Vishwadeep Khatri I like Rohit's comment above where he says ethics and CSR should come into play to ensure that cheating of either kind does not happen. Vishwadeep Khatri There is another angle to this - If you ask the customers what will they prefer - A car that is easy to maneuver, high on fuel efficiency, and more convenient OR a car that does not pollute, what do you think they will say? If most customers prefer the former option, an organization may decide to delight these customers while ignoring long-term human good/ ethics. Today after the news, if Volkswagen sales continues to grow, and customers (even after being fully aware of the cheating) do not bother to check if the emission control is tampered, is it partly the fault of customers or do you blame the company squarely for it? Vinothkumar C Legal Cheating will always be present due to the presence of Stock Market and quarterly results. This is where CAVEAT EMPTOR (Customers Beware) comes into picture.... Illegal Cheating is beyond the limit. Playing with big risk. Just one incident dropped Volkswagen from 2 to 4. Recovering this would be big task... Steve Borris Consultant and Author I think we sometimes tend to forget that the purpose of business is making a profit. Whether we like it or not, that is the way it is. For example, countries set their own tax laws, which leads some to set up in those with the lowest tax rates. What this can mean is that companies in one country can pay less tax by (as I understand it) having headquarters in another. This, being the case, we need to ask if the company, by obeying the laws in the countries in which they operate, are doing something wrong? I kind of think not. This creates the debate on "tax avoidance" verses "tax evasion". Only one is is illegal but we can debate if they both are, or are not, acceptable - and yet most of the public have no idea that it is even happening unless someone else highlights it as an issue. So, can we have the same situation in production? Of course we can. The manufacturer is obeying the letter of the law, is he not? But, as a customer, are MY needs being met? In lean, we recognize the need to provide customer value. So, if we promise "x" but provide something different "in normal use conditions" are we meeting the customer's values? The doubt has been raised and then the question has to become: who can we trust - legal or not? Trust has now become a major issue. How does the consumer know if anything he buys is represented by the specifications in "real" use or a lab - and what the difference is, if any? Can he ask the salesman is there are any special steps taken to optimize results? Would the salesman even know? It will be interesting to see if we adopt an attitude of appeasement ---- (...if one does it they probably all do...) and carry on as normal. Or, just maybe, the regulations will be changed to reflect (what I, and probably most folk would see as) the "spirit" of the laws. Steve Manish Sehgal I agree that main motive of doing business is profit. No business can survive without making money but trust is also important for business sustainability in long run. Cheating whether legal or illegal breaks the consumer trust whenever it is discovered. This causes a huge dent in brand value and it takes a very long time to recover the same. The cost associated with dent in brand value is huge and can wipe off all the money made in terms of profit. It is time for consumers to wake up and make these companies realize the importance of business ethics. Donald Kerr I wish cheating had bigger longer ranging repercussions but some industries (e.g. cigarette manufacturers) or others (Goldman Sachs designing products to fail to recoup money on "insurance") seem to prove that people's memories are short. Steve Borris Consultant and Author Donald Legally, cheats that get caught will face changes. But people have short memories for a lot of things - particularly politics. Maybe social media will act as a reminder. Steve Shamik Kumar interesting discussion, no doubt that business is for making profit but there is something called corporate ethics ....I also understand customers expectation is top priority(lean, ...) but then customer is not always aware of repercussion of its expectation, e.g. in pharma people don't always like bitter taste of medicines but then it doesn't mean manufacturer will start adding sugar to meet the customer requirement and making it tasty and this is where regulatory/ statutory/ legal bodies come in picture to ensure customer is safe, even if it is not aware. In case of VW the impact is not direct rather indirect and that's unfortunate we tend to ignore the most important aspect of our survival i.e. environment. Peter Martin It would probably be more of an issue if it was only VW doing the cheating. In reality it is only VW that has been caught. I suspect all the main manufacturers are cheating in one way or another. All manufacturers quote very similar emission and fuel consumption figures for similar cars so if one is cheating to achieve this then most likely they all are. Just unfortunate that VW was the first to get caught. This is a perfect example how possibly unrealistic targets can drive the wrong kind of behavior which is something we all see everyday in the workplace. Targets and incentives can work well but can also cause a whole host of unintended consequences. Malcolm Campbell Ethics sits above the law but only the law can be enforced. (unless there is massive social re-vault) As corporate pressures are sometimes stronger than an individual's good will, the intent of the law will never be any better than the letter of the law. Unless we invent an entirely new legal system Bernard Brinkley Well said, Malcolm. Vishwadeep Khatri "Intent of the law will not be any better than the letter of the law" - Terrific quote, Malcolm. You mean to say that intention element of the law that does not have teeth is worthless. "Corporate pressures are stronger many times stronger than individual goodwill" To summarize our discussion so far - Corporate pressures can lead to cheating and cheats who get caught have to pay/ change. Cheating is widespread and will most likely continue unless social forces group themselves against it. Legal and illegal are just terms - smart companies will make cheating look legal. It does not seem that society is likely to stand up in a massive campaign against such an issue. Does this mean companies that do not resolve to cheating will fall back in the numbers game sooner or later? Are we trying to say that the pathway that lead to top positions in competitive business inevitably move through the lanes of cheating? Malcolm Campbell Thank you Vishwadeep. Please send royalties for the use of the quote to..... I think the risk is exactly as you say. Corporations tread a fine line to win business. What is over the line for some is not for others. Those that do not consider it over the line (or do not care) will gain a competitive advantage unless a court stops them.... Sadly this sounds like a need for more regulation. How do cultures that do not value the written word so dearly manage this type of issue? Sunderesh Udayashankar The issue is also about legality along with morality. There could be many circumstances where something is not legally violating, however, morally it could be devastating. so companies have to stand up to their ethics and business conduct and demonstrate that they are morally binding rather than being legally correct. Sujan Chakraborty I do agree with Sundaresh. It actually comes down to Ethics... which can be defined many ways but I think it is "Obedience to the Unenforceable" in this context. However it wouldn't be out of place to ask, is the "Value" well defined? What matters to customer is what happens on road, not at test conditions. Hence, should there be a legal mandate for companies to declare a max. level of emission on road? Akarsh B N - PMP®, PMI-RMP®, RABQSA CSSBB hello. I have not come across anything called legal cheating. it can only be called innovation if organizations play with loopholes of the legal framework. however, in the case of VW, it is more of going against the regulations intentionally, than bending it and also charging customers a premium for it without the customers' knowledge. Business at the cost of ethics (knowing that they are cheating but keeping quiet about it) is very dangerous in the long term - for the trust, brand name and business itself. Finally, it is also political in the sense that this report was apparently out in public domain more than a year ago but is being raked up now in the wake of upcoming elections in the US. Beverly Daniels As Sujan alludes to, the dilemma we face can often be in how we enforce the 'law'. For emissions - or any test for that matter - we must remember that usually only one vehicle is tested. Or at the most a very small sample size. So we are really only getting a single value; as all quality professionals know a single unit cannot explain the total variation in the population. Are these randomly selected units? Or specially selected and submitted to the testing agency? who decides this? And how is the test conducted? under what conditions? Certainly it isn't conducted across the full range of use conditions...The test is usually highly specified and so the test itself accounts for the difference between the published test results and what the driving universe experiences. Hence the phrase, "you're mileage may vary"... is this 'legal cheating'? Well the automakers could very well take advantage of the test method. As much as we are going to be the victims of the test method's inability to cover the full range of variation in any given design. As always these things are far more complicated than they first appear.
  3. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Vishwadeep Khatri There are situations where data is hard to collect or the occurrences are too infrequent. Highly experienced managers, however, are happy providing their guess about data through rating scales. They believe that the ratings are pretty close to real data. As a problem solver, would you go with such ratings to find causes? Partho Banerjee LSSBB I don't think that ratings alone would be sufficient. In such cases. I feel a further drill down is required. Besides we can look out for measurables which are closely related (pre or post stage) to get some idea of the affected process. Shamik Kumar I understand that "not using human intelligence/ experience" is one of the wastes in lean. So we need to appreciate people having experience and providing feedback based on that. Now as a problem solver i would rather prefer to take their experience into consideration and try to validate it, if possible, than to simply ignore or believe it. Mayank Gupta I have done this for one of my projects with pretty good results. This is what we did * Validated the domain expertise of the agent providing the ratings * Took ratings for all the potential Xs where data was not available (prevented the bias towards a particular X) * Cross verified the ratings with a focused group So as a problem solver, I would go with such ratings, but one needs to proceed with caution and there are additional checks and balances that need to be taken care off. Manoj Singh In absence of data always trust your experiential logical regression ( domain knowledge). It really works Satish Karivedha When data is not available to understand root causes, then we relying on prediction theory. There is nothing wrong in going with ratings. While considering rating, ideally one should look at inputs coming in from SME's. Secondly larger samples will give good insights. In this entire context of findings, degree of risk is relatively is on the higher side as there is no data to prove. We are going with assumptions. Partho Banerjee LSSBB I agree with Satish and Manoj but what would be our approach to get more sampLEs as in this scenario the rating and that too from those who belong to the function or are affected by it(stakeholders) since they have already given their ratiNg. Who else should we look forward to for our study. Beverly Daniels If you don't have any data how do you know you improved the performance? I worked on a Problem that resisted 'solutions' for over 20 years. People guessed, used ranking, fishbone diagrams and voted on the most likely solution, called in subject matter experts, etc. None of the actions made the performance better. They didn't use any data. We got data. It took a bit of time, but we baselined the failure (Define phase), Determined that we could in fact measure it with some repeatability (Measure), ran a couple of experiments to recreate the failure in-house and understand the causal mechanism (Analyze), determined a solution that we proved with data would eliminate the failure mode (Improve), then added a set of controls on the critical inputs and the output performance to ensure that it didnt' re-occur. It took time to get the data but no where near 20 years of failed guessing without data. Now I'm not against the quick fix based on past experience or subject matter expertise as long as it's easy, cheap and quick to implement and I'll know rather quickly if the solution worked, but we need data for that last part don't' we?
  4. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Gorur Sridhar Some are of the opinion that a problem can have more than 1 root cause, but I feel it negates the very definition of ROOT CAUSE i.e. Root of the causes which is only one. Once you solve it which is a major contribution factor then the second most contributor can become the root cause to be eliminated and so on. Vishwadeep Khatri Interesting question, Gorur. Working on one factor at a time has its advantages in some cases especially if causes are not independent of each other. I do not think I would limit myself to one root cause at a time in a team problem solving situation, especially when members have capability to work on few independent issues simultaneously. What do you think? Gorur Sridhar That is OK ..working in parallel, but what I mean is that, definition of root cause is that there is only 1 major cause to the problem. Once you address the next most significant cause becomes the 2nd root cause hope this is right. Vishwadeep Khatri Let us talk about delay in flight take off as an example . I may find baggage loading delay, delay in food supply and passenger security check delay as three reasons having same frequencies. It may not be easy for me to consider one of the three as a single root cause or major contributor. Considering all three as root causes should be fine. What do you think? Gorur Sridhar We have to try to give weightages for the factors either in terms of loss of revenue or quality or availability etc. so that we single out one of them and then proceed. Mohan P B Would agree with VK in that limiting oneself to a single root cause may be detrimental to solving the problem completely. The singular aspect of "Root Cause" is figurative and need not be taken literally. In most real problems there are many root causes, resolving which would solve the problem completely. Resolving only one root cause will improve the situation but not completely. For example, if the problem requires improving Quality from (say) 90% to (say) 99%, there could be four or five different root causes, resolving each of which may improve the Quality by 1% to 3%. But resolving all of them will help us reach the target of 99% Vishwadeep Khatri I agree Partho, it may be a system of interacting root causes in some cases ( where interactions exist ). The focus in such cases shall be on the system and not one specific cause). Gorur Sridhar That is ok, what I am trying to understand is that is it Root Cause or Root CauseS? We always ask to find out the root cause and finding out the root causes does not sound to be logical! Partho Banerjee LSSBB I agree gorur, the terms stands as "Root Cause" it is more of a symbolic term. generally we reffer to as "Finding the Root cause" because, ultimately we put 90% of our efforts in diminishing the effects of that one main cause and rest for the interacting causes. The ratio would vary depending on situation & based on the conclusions from the data. Shamik Kumar It is one of the most common errors in problem solving that we make, with assumption that for any problem there will be only 1 cause which can be called as root cause, its important for team investigating the problem to keep an open mind and analyse each probable cause and their effect thoroughly. Root cause can be one specific leading to your problem or it may be because of interaction of multiple causes, in second case it will be difficult to identify one problem item, instead the team should work systematically to validate and correct all identified causes, also in second case some time people do call it as system failure/ failure of risk assessment or etc so to just arrive at one problem but then it becomes a generic cause and most of the time they end up missing on some or the other smaller problems which in future becomes big. Mayank Gupta Gorur - a problem can definitely have more than 1 root cause (or "Root Causes"). I'm surprised to know that there are people who think that there has to be only 1 root cause. How many of such causes you deal with at the same time depends on the appetite of the business and the project team. Christopher Sequeira (LION 2900+) It is good to find out the root causes and their interconnections. Figure out which of them is the highest contributor and attack it before the others. Doug Ford Problems can absolutely have more than one root cause and usually do. The trick is to understand the tree of the problem to see the various cause and effect branches and understand the relationship/dependencies. Once you can understand that, you can figure out the most effective way(s) to break the cycle to eliminate the problem from resurfacing. Ashok Motwani Gorur In your question you are suggesting that perato thinking is utilized where the biggest bar (which contribute most to the problem) is tackled first. Correct me if you think I am off-base. However others responses are also correct where independently no one cause could improve the situation at hand but interactions can be the prime objective to handle so problem can be solved. Picture how "Root" looks like. The answer is - it depends. From the trunk of a tree or plant many shoots or one shoot starts and then divides into many branches. Similarly it is dependent on the type of problems being solved (maybe.....) and the perspective of an individual/team who is leading the problem solving event. Question or comments are welcomed. Ashok John Predmore I think a problem can have more than one root, same as a plant can have more than one root branch. I do think some people mislabel every factor in the causal sequence a root cause and that is a disservice. My definition of root cause is 1) something that is not supposed to happen and 2) something we can control or prevent. By this definition, if I slip on an icy walkway, the root cause is not the ice. If it is raining and the temperature is below freezing, ice will form. I cannot prevent the rain or the cold temperature. In this circumstance, I need to probe deeper using Five Whys, for example. Why did I fall? Maybe I did not know the temperature. I can hang a thermometer by the door. Maybe I did not see the ice. Improving lighting could help. Maybe my shoes have poor traction. Better shoes could help. Any one of these could be considered a root cause as long as fixing one of them is sufficient to prevent recurrence. Ashok Motwani John Good example and explanation. Thanks Christopher Ayres A root cause it the deepest cost of a concern, kind of like shedding the other concerns to get to the root. So the answer is one. Christopher Ayres Miss understood the question. Christopher Ayres At the same time there is only one critical x. Take it easy on me I just got verified green belt. Sharon Vollers Well Root cause theory generally teaches that there is only one root cause to a situation and generally I support that position and believe we should push for that. Having said that however, Absolutes tend to be problematic in that they represent our current theory/ knowledge and close us off to the possibility of other discovery. This would bear some scientific analysis. Raju MRC "Problem" is the overall high level dissatisfaction/ non-compliance that is observed in a system. Every problem with have one or more "Failure" associated which together lead to the problem. Each "Failure" in turn will have one or more "Defect" associated due to which the failure occurred. Root cause is always associated with the "Defect" - We do the analysis at this level to know, what led to defect getting introduced?. This question always leads us to many causes, but only one "Root Cause". So, when we look for root causes from a problem perspective, there will be many. Martin Steimle Hi, during root cause analysis you want to identify "the" root cause and eliminate it. Some people use the word "true" root cause. One of Toyota's principles is to identify the root cause to ask the 5 Why's which seems an efficient way to come close to "the" RC. During this journey of asking the 5 Why's you might face several branches and you have to decide within the xfct team which branch to explore next and then do experiments to justify the identified RC. hope this helps a bit, Martin Pascal Chaloyard If you are considering one problem usually you have combination of a risk factor and a triggering element Steve Borris Consultant and Author There can be more than one cause for an issue. The point about the Pareto made above suits the situation well. I do get a bit worried by the belief by some problem solvers that the right solution is found every time. This is simply untrue. I would very rarely recommend implementing multiple "solutions" at once. I have seen situations where multiple improvements were made simultaneously and, as well as the problem not being fixed, new problems were created. But I guess it depends on the complexity of the solutions. Just because an orchestra has a lot of instruments does not mean that they should all be playing different tunes at the same time. Identify the possible causes. Identify the solutions. Define monitoring tests to ensure resolution. Prioritize for implementation and plan a schedule. Only if we can guarantee that implementing a second solution will not have any negative consequences or if the problem area is very hard to access, would I consider multiple fixes --- and I would need to be sure all the folk involved were fully aware of the risks. Steve Sanjay Rawat (1500+) As per my opinion 5 why analysis is for finding the ultimate root cause of the problem.. There are possibilities that we might find minor causes around the main cause which we can ignore and continue to identify the main root cause. Martin Leighfield Prof Dip Mgmt (Open) Using the example above, is the level of dissatisfaction the problem or is it the effect of the problem? People leaving could be the effect of bad customer service so are we looking at improving the level of dissatisfaction or are we focusing on the bad customer service. And focusing on one cause may not be the true answer as the output of a process is the sum of ALL it's inputs. Ashok Motwani Steve, well said about how to implement countermeasure. No cause should be considered insignificant and each countermeasure implemented could create another symptom. Martin, you are right about understanding first which problem is being solved so first step is to clearly and succinctly definition of Problem in measurable units must be completed which will provide the scope and time frame. Once a countermeasure achieves the quality levels/ goals required or expected by the customer (internal or external) there comes a point in time when additional analysis and finding causes becomes unhealthy financially and becomes a drag on resource availability. Comments are appreciated. Ashok Partho Banerjee LSSBB I would like to add to Ashok and steve..for the same reason we identity our secondary parameter/s which shouD not be affected while we are workng On the root cause/s. So active targEting of primary parameters and monitoring of secondary parametERs Alan Charles Some great thoughts, hope I can add by coming at it differently Consider every problem with a make ( why it happened) and a flow ( why it got out) you will 2 root causes from these. Based on whatever you call it, FMEA, QA Network ( Toyota speak) and several versions/themes around it, you can have a strong c/m to RC1 and no need to have a RC2 c/m pending a matrix evaluation. A bit off topic that bit, but its a way to show you can have multiple RC's. They are all important bits, but the top of the funnel, to clarify the problem and then cause and effect an have a massive bearing on the 5Y you go down. Depending on anyone's level in a company you can ask another Y, RC's on systems are the key. Ignoring all this, and may seem strange, but what did the people get out of the investigation, this shows training, culture and development more than the RC to some extent. Its the ownership of this element by the responsible section which is often lacking leading to RC problems. Femi Obiomah While initial theories have stated that there should be one root cause, I believe like in other cases, modifications can be made drawing from experiences and applications. We now have the fault tree as well. very few trees have one root. Take customer dissatisfaction for example, its easy to attribute poor business plan as the root cause. There are cases however in which even when there is a proper business plan you still have customer dissatisfaction, There is the need to address all the possible causes. Steve Hall As a minimum, each experienced issue has 3 root causes. Occur - why it happened in this case Escape - why it passed the process and on to the next customer Systemic - what element of failure avoidance failed to predict and manage the issue There can, of course, be multiple root causes of the occur. Sometimes complex issues actually interact, and the whole issue might not be an issue, unless all the root causes happen. We don't have to hang on to the concept of having a single "root cause" if it does not fit our situation. There could be a pareto of issues and a significance of contribution - as can be found during a Design of Experiments event (for example). Escape is usually simpler, in that the issue was not being either looked for or detected - which brings us to the Systemic root cause. Was the issue, or its significance, predicted / predicted correctly in the failure mode avoidance activity? This can be done during D7, but it helps to start the thinking off, when working through the Root cause analysis. You could argue that this systemic root cause, is the truest root cause of all - which I think it is. So, in conclusion, there could be many root causes, that all happen at once, or individually. Verification and Validation will go a long way to prove that you have made a positive impact on the issue. But, it might not pick up everything in the first round of the issue, but you might reduce it to acceptable levels - however that acceptable level is defined. Steve Hall BTW - complex issues with potentially multiple root causes, is where the world of 6-Sigma takes us beyond the discrete, time line/event orientated world of 8D Ashok Motwani Thanks Steve for bringing up the 3 points of failure in a out of standard condition which is called "defect" for which root cause analysis has to be done. Typically an individual or a team focus on one of the failure mode and are unable to eliminate and prevent the recurrence. Great response to the original question! Mark Reinard Sarsonas I'd say that'd depend on the scope and limitation of the charter. The more specific it is, the greater the chance we can point to a single root cause. I haven't done a project to support this idea though. Christopher Vallee As a TapRooT Root Cause Instructor, we started our process on the concept that multiple roots can cause individual problems. These individual problems (actions or actions) either initiated the incident (the worst consequence), failed to catch it in time to mitigate or made it worse. Each one of these problems having their own root causes. Think about a fire or explosion. It takes at least 3 main ingredients to come together in the right sequence to create it. We can argue all day like the number of disagreements in the posts above and been no closer to preventing the incident from occurring than we are now on deciding on one for multiple root causes. There is not one root cause for anything. There are instead multiple issues that if we reduce or eliminate will reduce the probability of an incident occurring again. Worst case scenario, lets say we voted on the fuel source only for the fire and then fix it this time. By ignoring the other factors, the uncontrolled ignition is just waiting for it's next fuel source.
  5. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Oscar Rodriguez-Gonzalez They are complementary Vishwadeep Khatri Hi Oscar, the question raises the concern that people consider lean to be merely equivalent to automation. Any thoughts? Manoj Singh Good automation is one part of sustainable Lean. For sustainable lean we must focus on PPT (people, process, technology). If only good automation = sustainable Lean then all Fortune 500 would have sustainable lean at the same level as Toyota. Partho Banerjee LSSBB I agree with Manoj. Automation is a Lean of the enabler tools. There are processes which cannot be operated unmanned and still can be called a sustainable lean process by using other tools such as Poka-yoke etc. Please correct me i am wrong Aniruddha Sahasrabudhe Hi Vishwadeep, Based on my real life experiences in Automobile & Aerospace Manufacturing & Engineering Design Services , my comments / views on "sustainable lean = good automation" are as under A) Primary Intent of a " LEAN " Process in an Organization : Primary intent of making a process " LEAN " is to eliminate waste ( of Men , Material , Money , Idle Time or any other activity that does not add any value to the process ) from it so as to make it Sustainable , Optimal and Deliver Defect Free Products / Service conforming to the Requirements / Specs.of Quality , Cost & Delivery Time of the End user / Customer. " Good Automation " : Frankly speaking I do not know definition & examples of Good , Better , Best or Bad Automation. I go by simple concept of Automation that primary intent of the Automation is to a) Reduce / Eliminate element of subjectivity due to Human fatigue from a repetitive task and ultimately deliver Defect Free Products / Services to Customer and Ensure uniformity of Quality of Products / Services delivered to a Customer Examples : i) Currency Counting Machines in banks : Here can't think of manual counting of currency deposited / withdrawn from a bank due to high volume & value of currency deposited / withdrawn each day. ii) Process of Painting of Cars - Here can't think of manual painting instead of Robots due to high volume of production of cars / day iii) Cold Drinks / Beverage Bottling Plants - Here can't think of manual filling & sealing due to very high volume of production of Cold Drinks / Beverages / day. To sum up I am of the opinion that * There may be " sustainable lean " processes with partial or without " Automation " in organizations who do business in a Service or Non Manufacturing sector. * The other possibility also can't be ruled out i.e. There may be some " Automated " processes in an entire supply chain in Manufacturing sector but the entire process per say may not qualify to be truly called " LEAN " since there may be scope for reduction / elimination of waste in some part of the process. Therefore one should not blindly state the Hypothesis "sustainable lean = good automation" without really defining / knowing the details / context of the process being discussed. Regards Aniruddha Vishwadeep Khatri Mr. Aniruddha, Good to see a well presented detailed comment from you, like always. Oscar Rodriguez-Gonzalez I agree with the comments. And to add a little more: Things that Automation does not have: * Lean-thinking * Lean-business Things that Lean does not have: * Robotics * Machine-learning Things that both have in common: * Error-proofing * Philosophy of waste reduction There is a lot of ground to cover in this discussion. Probably the most illustrative example to visualize their compatibility is the use of Agile (the IT version of lean) to create software (automate). In this case lean (agile) does the project management and automation is the product. A Value stream map could also help to identify the areas to be automated (improved). Probir Bosé, MBA, LSSGB Aniruddha - I really like how your response is so brief and succulent. Actually, I was a bit amused and did manage a slight chuckle when I read it !! You see, first I read Vishwadeep's comment about your post which was on the lines of "a well presented detailed comment” and was slightly intrigued. Needless to say, when I clicked on the link to read further, I was actually visualising your post to be very literary, with detailed explanations on different approaches and conceptual theories and frameworks. blah, blah..; instead I was gently reminded yet again that often even the most simplified and 'to the point' observations can be just as remarkably effective. Hehe..Like. Uma Great read Aniruddha! well explained with examples!....While I second you that a direct equation could not be brought between sustainable lean and automation without the context (best example is healthcare/hospitals where this theory is not proved), can we say an intelligent automation (not only to replace manual effort, but also eliminates the NVAs) accelerates the lean methodologies and sustain process excellence. In these many years, the iterations of lean concept has evolved to include definitely a level of technology integration. My 2 cents! Siddarth Shankar Instead of calling it automation ( which is fully automating, and not realizing the risks behind building a monster) I would follow Autonomation which is more an intelligent automation. Helps to have a bit of manual control instead of e2e automation which is not controlled. Information / material flow with out manual touch is automation , information / material flow with a manual touch only on quality controls is lean. Cause automation is easy when it come to standard process. The challenge is when you have to bake in all exceptions/ deviations customer wants and then lean comes into play to decide what really is value add and what's not and what needed but not value add. Sivakumar Viswanathan Needs to be viewed in a more flexible context. Lean is about building a culture of questioning status-quo, by bringing process improvements at the grassroots level. It involves a careful and smart mix of automation with human touch - autonomation !!! Hence automation is part of the game, but if there is a better way of performing a task(or process), then automation should be adaptable.....!!! Venkat Narayanan Iakshminarayanan No. Even many companies are doing automation activities separately from lean they show it as kaizen. Lean to be looked at holistic approach Gopal Bhat I agree to some extent. Actually LEAN principles are substitute to automation where practically full automation is not possible. However, the big difference between LEAN and process automation is that LEAN leaves further scope for continual improvement where as automation in principle puts an end to CI thinking. Surendra Patil Hi Vishadeep, When you posed this question for discussion, I hope you want us to discuss the concepts behind the basic tools of Lean and surfacing out the misconception of relating part of tools and technique referred as whole concept. Most of people have already explained in above discussion about the definitions of both ‘Automation & Lean’ Most of the time in organization mangers are tend to use the tools & technique in which they are good, (if all you have is a hammer, everything look like a nail) undermining the effect of it on whole organization as a system. When Taiichi Ohno was first used the basic principal of maximization of flow of product and developed techniques to solve the problems in his own company, and never referred as Lean Management. But it was named as ‘Toyota Production Systems’ as because it was developed to suit to certain environment. Later on which was named as JIT ( just in time ) and Lean management by team of professors under leadership of prof. James P. Womack those who have visited Japan for study the success of Toyota. While applying any tools & technique of lean management in any part of organization one should not ignore it’s effect on total system and not on part of the system in isolation. Most of the time the improvements done on department level (low level automation is a good example) dose not lead to organizations performance. As because the underlying assumption that sum of improvements at local level is equal to global improvement, is not valid. We should consider the organization as system of dependent events and will carry the effect only at leveraging point. Hence automation in part of the system may or may not have direct effect on organization’s performance. If it is downstream of constraint then it will lead to increase the inventory and if it is upstream then will starve for material.(Assuming the automation will increase the rate of production.) If automation is for reduce the fatigue / human error and increase the quality then certainly it will add the value. Effect of automation can also be verified with financial measures, which will link to organizational performance. While implementing Lean management thrust should be given on flow maximization and not on the reduction of cost with using tools and techniques. We particularly Indian managers are wise enough to articulate the solution first and then fixing it in a particular tool & technique. (this may call for separate debate).
  6. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Vishwadeep Khatri As a team leader or coach, how do you use questions to help guide your team towards a meaningful direction? Is it always better than providing a solution? Narender Sharma Questions always make a map of the subject in the mind of learner. Hence make him curious to know the answer, here when trainer or coach explains he grabs maximum. Steve Borris Consultant and Author I like to encourage folk to question everything. Not everyone, however, likes to be questioned. I see the question as the first step in the improvement process. It does not need to asked aloud - but often discussion adds clarity - it can simply be a trigger. It also helps to set the background to a situation should you be offering a new process and avoids the "Why are wanting to do this?" Steve Steve Borris Consultant and Author I hope to follow this thread. Need to find a way to do so, now that LinkedIn no longer send updates on new comments. Indeed their decision leads to a lot of questions... Ste Vishwadeep Khatri HI Steve, I just realized this - Linkedin does not seem to provide updates on participated discussions any more. This is going to make life difficult for regular users. That is a really confusing change they made. Satish Karivedha If you are in the consulting side, to provide a solution to the business need, one need to ask required questions to understand AS'-IS state, before providing any solution. This gets connects to Demings Profound knowledge which was being explained by Deming with four elements a) Appreciate the system understand the variation c) factors influencing the system d) Use theory of knowledge thru SME's. By using Deming's profound knowledge ask right questions and provide right solution to the business need. Mishtert T As a team leader or coach, asking leading questions would help the opposite party to realize where the gap was and how a situation could have been handled differently or better, which will help in retention as well as replication in future.
  7. .demo { border:1px solid #C0C0C0; border-collapse:collapse; padding:5px; } .demo th { border:1px solid #C0C0C0; padding:5px; } .demo td { border:1px solid #C0C0C0; padding:5px; } Vishwadeep Khatri In another discussion, someone commented that a process is never extraordinary. It is just good or bad. What do you think? Aniruddha Sahasrabudhe Answer to this question lies in precisely " Defining " all the Inputs , nature of Interactions e. g. Simple or Complex between misc. activities in the " Process Blackbox " & Quality of Final Deliverables ( Intented as per some Customer spec. v/s What is Actually delivered ). For example Process of Criminal Investigation with global ramifications may be called complex and extraordinary whereas most of the established & known engineeing processes in manufacturing & service sector may be called as just good or bad. Vishwadeep Khatri I get your point of view. There are many ways of looking at processes. I am not in favor of just a good/bad assessment. Why not an ordinal scale? Most certifications including ISO 9001, SEI-CMM, COPC, TL 9000, AS 9000, ISO 14001, ISO 22000, HACCP, etc have a process approach. When the assessment is done, would you just say that the process met expectation or did not meet expectation? Or would you consider some processes really far ahead of what is fundamentally expected of them? Uma I restrict being binary by saying a process good or bad in achieving the results...defining such adjectives (extraordinary, remarkable, incredible) to process is an herculian task and to quantify as well! As Aniruddha stated there could be complex process depending on the sector and its limitations, but a complex one need not be an extraordinary process! Vishwadeep Khatri Customer satisfaction assessment used to be a 3 point scale in hospitality industry. Below expectation, met expectation, above expectation. Now it has moved to 10 point scales at most places. Vishwadeep Khatri Also, Process maturity assessment (CMM or others) is generally a five level assessment. I like the five levels approach quite a lot. One process may be at a level which barely meets the requirement. Another may have controls on all key inputs and flexibility built-in to respond to changes in supplies, market or competition. I would not like to consider them as equal in my assessment. Aniruddha Sahasrabudhe In my opinion we must also look at & link Customer's feedback with " How& What the Process ultimately Delivers " to the Customer. I will give a simple example of smooth landing which all of us must have experienced over hundred times by now. Depending on air traffic over an airfield , ATC gives clearance to a Pilot to land on a particular runway. As all of us know each runway has certain width and there is a white line right in the middle of runway which the Pilot is supposed to follow for a safe landing. Now imagine for sake of understanding that this line as the spatial reference ( 0,0,0) and the width of runway as the Upper & Lower Limits. A Pilot adheres to the ATC's landing instruction & most of the times lands safely.But imagine what would be the mental & physical condition of Customers ( passengers ) if the plane lands just couple of feet close to either side of the runway and then taxis to the park bay.Theoretically Pilot is still well within the width of the runway , the aircraft is also intact but some of the passengers might faint. In this simple example KPIs ( Key Process Indicators ) of Customer's satisfaction are are Quality of Inflight service , Operational Punctuality ( Actual Takeoff & Landing time vis a vis , that mentioned on Flight schedule ) , Safety & Passenger comfort. The point I want to stress is that there may be several ways of looking at a process by measuring several parameters and applying multiple criterias but we must always remember that ultimately the deliverables -- Product / Service must comply with Customer's requirement and it is the Customer's choice / prerogative to call a Process as Good / Bad. Any Process certainly can be improved by very well established methods of Process improvement which are beyond the scope of current discussion hence I am not going into those details.
  8. Shahjahan H I know, Data analysis is all about analysing the big junk of data and presenting it in a meaningful way whereas 6 Sigma is all about continuous process improvement. Any other thoughts between Data Analysis Vs Six Sigma? Vishwadeep Khatri Hi Shahjahan, great question. Six Sigma became popular because of its focus on data driven decision making. There are decisions to be made in each phase of a Six Sigma project and we prefer to make those decisions utilizing the right kind of historical, experimental or simulated data. So, data analysis (before analysis we need to ensure that it is the right data for the right purpose in the right format and being planned for use for a fitting purpose) is the core part of Six Sigma You are likely to analyze some kind of data in each phase of a Six Sigma with right validations and a clear objective in mind. Six Sigma is likely to provide good direction and more sting to data analysis. Shahjahan H Thanks VK, I understand about the 6 Sigma - step by step approach (DMAIC / DMADV). Can we say 6 Sigma is the base for Data Analysis / Data Science? As, both 6 Sigma and Data Science uses Statistics techniques to a large extent. Also, Data Science is the advancement of technology, where the simple Excel and Minitab can not be used extensively for Data Science and it requires tools like R, SAS & Python for extraction of the data and to make more sense of the data? Shahjahan H I am coming to the point where, In software industry, we say "Mainframe technology" is becoming outdated / old one and the companies are moving towards advance technology like SAP/Oracle/Cloud etc and MF is considered as Legacy system and slowly degrading. Similarly, at some point of time, will 6 Sigma be considered like that and more preference will be given to Data Science professionals (due to the advancement in technology) and what will be the future for 6 Sigma BB / GB / MBB professionals? Vishwadeep Khatri Hi Shahjahan, you have brought out a very valid point. Thanks for bringing that up. The way we look at Six Sigma today is very different from how it was started by Motorola in 1987. It has been evolving. An MBB curriculum now a days has included Simulation, Advanced DFSS methods, Creativity and Innovation etc. Current outlook - Even today, for a Lean Six Sigma project requiring Big Data analytics, we ensure that experts in that area are part of Lean Six Sigma team in the right phases of the project. Future Outlook - If there is enough demand about Data Science including Big Data, Master Black Belt curriculum is likely to be revised to include these. We carry out surveys every year to be able to get recommendations in for revisions in MBB curriculum of our own. Shahjahan H Thanks VK for your time for giving the clarification. Soon, we can expect a change in the MBB curriculum and we always look for a betterment of the course and eager to learn the new methods of learning with the technology change. Sushant Kaul Valid point : Lets assume a study ; there must be at least 1000 Indian Million revenue based companies. It's not possible to be without a huge MIS Or ERP system. Tons of data &graphs !!! So if data & graphs are already available then are all 1000 companies making best out of available resources !!!! As already highlighted DMAIC is one of the best approach to deal with data ... Paul Astle Shahjahan, as you know the success of LSS relies on fact based decision making. You will always need to do some sort of analysis on data that you have either obtained from a data repository or collected yourself and verified the integrity of. If you personally don't have the necessary skills to be able to analyse your data set, then you need to ensure that you have someone on your LSS project team who can. As a MBB you are expected to have an advanced level of statistical analysis skills to accept/reject the null hypothesis, ensure correct sample size and test for statistical significance etc. So, as I see it, if your data set is that large or complex that you need the expertise of a Data Scientist and an advanced at rest analytics platform to undertake the analysis, then this is outside of the MBB's remit as it is a specialist/full time job in its own right. The MBB should however be able to recognise when the analysis is beyond them and as I said earlier, draft the appropriate colleagues with the right skill sets into the project team. Beverly DanielsAnother way to look at this is that "big data" and its analysis is a tool in the six sigma tool box. It can help us get to the causal mechanism (but will never totally replace hands on experiments) and it can provide us with a control mechanism. But it won't replace MSAs and it can't help us develop, validate and implement solutions (except in a very limited and special way) Shahjahan H Hi Beverly, do you think 6 Sigma expert is really required for the Data Science / Big Data Analysis? Statistical knowledge + Tool knowledge (R, SAS, etc) should suffice, what you say? Rip Stauffer Statistical tools are just some of the tools in a BB/MBB's toolbox. Half of Six Sigma (and the harder half, for a lot of people) is the qualitative stuff...process knowledge. Statistical theory and data analysis certainly exist in many places outside of any Six Sigma framework. Six Sigma is just one scientific approach to continual process improvement, that happens to have some statistical tools attached. Vishwadeep Khatri Hi Shahjahan, let us go deeper into this. What is the objective of data science in a company. I have seen several data scientists analysing data and providing inferences that no one utilizes. I have also seen data scientists contributing very well when they work with a clear objective within a lean six sigma project team. Shahjahan H This makes sense - Working with a clear objective (understanding and achieving business/customer objective using DMAIC approach), where 6 Sigma professionals are trained with. Thanks Sushant, Paul, Beverly, Rip and VK for making this discussion more meaningful and got a better understanding of this topic. Beverly Daniels Shahjahan: Certainly there are many uses of Big Data beyond solving Problems and reducing variation within the DMAIC or Six Sigma frame work. In fact much of the early work with Big Data was not Problem Solving but forecasting and modeling of consumer behaviors and bio-health markers etc. ‘Black Belts’ are not typically trained – or skilled - for this application of science and statistics. Forecasting and modeling require a completely different set of statistical tools. We don’t need t-tests, confidence intervals and ANOVA tables for Big Data (we don’t need them for Six Sigma either but that is a different discussion). We need things like Principal Components analysis, multi factor Correlation studies, cluster analysis, etc. Beverly Daniels On the other hand there are few people who are skilled in “data science” or “statistical engineering”. There are many statisticians and many scientists trying to do this work, but what we really need here are a combination of science (or subject matter experts) and statisticians who first understand the difference between statistical analysis on Sparse Data vs. Big Data. (For example, Big data will generate a lot of correlations with p values <.05 simply by chance and thus this traditional approach has little relevance for Big Data). We must remember that “statistics without science is gambling and science without statistics is psychics”.
  9. Vishwadeep Khatri Well, it all depends on how risky the error is. If Type II error is much more riskier of the two (Type I and type II) and you want to make sure sample size does not become too large, you can decide to go for higher alpha value (while keeping beta low). If it is easy to get large samples, you can choose to keep both errors low. Ashok Motwani One can use this example to explain when the risk can be taken by increasing alpha value. If you’re analyzing airplane engine failures, one must lower the probability of making a wrong decision and use a smaller value of alpha. On the other hand, if you're making paper airplanes, one might be willing to increase value of alpha and accept the higher risk of making the wrong decision. Comments are appreciated. Vishwadeep Khatri Well, it depends on the objective. An aircraft's supplier quality assurance team may go for a high alpha but low beta while doing inspections with limited samples for supplier approval. This is so because a bad part being accepted is riskier than a good part being rejected. Ashok Motwani Dear Vishwadeep, I completely agree with your response. My example was for comparison purpose..while using 2 different scenarios using airplane as the product. Comments are appreciated. Satish Karivedha Alpha, Beta which are also called as type I and Type II errors, and they are inversely proportional to each other. When you are looking high degree accuracy of Type 1 Error, you are giving good room for Type II error. Coming back to the question when to use very low value of alpha and high value alpha refers to your need of study. Low value of alpha means, you’re ok with high type II errors, high value of alpha means, your focus is more towards controlling type II errors. Practically both of them are errors, based your requirement study your trading off one of them. These errors can be controlled by increasing the sample size. Ashok Sharma Thank you sir... Chuck Gillis The alpha value can be set for any value, it all depends upon your risk tolerance. Sometimes we'll set it to 20% when looking for directional signals..other times we will set lower when it absolutely has to be spot on.. a minimal risk scenario. I always come back to the adage of how much risk are you willing to accept. Sushant Kaul Hi need one input regarding Minitab 16. I generally report few kpi's using run charts. Every month I have to manually update the data base to have desired result. Even though graphs & worksheet is saved under project format I can't just refresh & update data in columns. Please share your inputs.. Erik Laufer The way I typically approach alpha is as a proxy for the type of analysis that I'm performing. If I'm in an exploratory mode, my alpha will be set higher. Perhaps 0.10-0.20. This makes sense when processes are variable, first dabbling with a process, we haven't utilized SPC/Multi-Vari-COV analysis/Gage R&R-MSA; or employing screening-type DOEs. More factors make it through the sluice, but I tighten that down as I move into characterization and optimization activities, ratcheting down the alpha to less than 0.05. I use the quote, "you will miss 100% of the putts you leave short..." Why incorrectly eliminate a possible factor by an overzealous alpha that is not warranted early on? The other graphic that you can use are the scales of risk...high impact of a mistake translating to loss of life, and/or, millions of dollars? Drive Alpha low...low impact, I can live with a higher likelihood of a mistake...let Alpha drift higher... Regards, Erik Jason Bodnar, Ph.D. Don't forget about using simulation theory to determine if the desired alpha is actually achievable given your situation. Think of alpha as desired versus computed. Vishwadeep Khatri Hi Jason, great comment! I am starting another discussion string on this - How to determine if the desired risk is achievable. Kiran Varri In a NASA situation a 1% alpha n in case of a extremely low risk scenario, a 20% alpha...n in between is why the world goes with 5%... Jason Bodnar, Ph.D. Commonly in engineering settings, the type 1 and 2 risks are prescribed in industry standards such as ISTM and others. Outside of an industry standard, type 1 and 2 are typically chosen as a function of the risk associated with acting on the statistical analysis results. This is the hard part. What is the impact of the observed Type 1 and 2 errors on the business? How risky or conservative can you be? Sadly, there is no single path here. This is the fun in understanding the mechanics of the statistical theory of hypothesis testing.
  10. Kaushik Jadhao Continuous Monitoring. Targets should aim for "best in class", with step wise improvements in processes, similarly new business requirements will define new CTQ's. Improvements or changes to processes should never cease. Craig Lester I agree with Kaushik Jadhao - Continuous monitoring however, I will add that you must remain engaged with the process stakeholders. This is a must. There should be no 'One and Done' efforts in anything - ever. The best systems and or processes will begin to degrade as soon as we step away. Vishwadeep Khatri Thanks for your responses. I would say - A dynamic monitoring system that gets stricter when process develops a slack and gets a bit linient when process performs excellently. When would you say is the right time to remove monitoring from a well performing process? Craig Lester I wouldn't completely remove the process performance monitor. I would tailor the monitor to key steps (pay points) of the process. Over a period of time, a process may degrade and if we do not keep our 'finger on the pulse' of the process, we may lose the gains that we worked so hard to achieve. PRASHANT VASHISTHA It has to be ingrained at the culture level. Develop a sense of ownership for the process. Am seeing this from leadership view and not operations view. PRASHANT VASHISTHA I agree with all my learned friends about the need for monitoring. Haresh Hiranandani I have a different view. Monitoring a process change for sustenance is a complete NVA as first time right principle should be built in the process change rather than post its implementation. Today we have NLP AI which learns and changes process as per previous learnings and ensures process improvement. Why would anyone want to sustain a process change when the dynamics around it are ever evolving, just a view. Craig Lester Haresh - For me, I was speaking to monitoring the process performance after the improvement change has been made. The improvement change may be related to more than just first time quality. When I establish pay point I use OEE such that I have an ongoing report card for process availability, first time quality, and productivity. While some processes have natural language processing (NLP) or artificial intelligence (AI) capabilities many do not and the cost to have this technology throughout the process (necessary if / when the constraint moves to another area) is often cost prohibitive. To your point - I agree that monitoring process performance as it relates to FTQ only would be NVA.
  11. Sahjahan H Outliers are nothing but the Special causes of the given data. We need to identify the special causes and remove them for the Data to become normal. Analysis with the normal data will always be more meaningful than the data with Outliers. Before removing any Outliers, we need to make sure that proper RCA is in place and actions are taken against them. Sometimes, removing one outlier will lead to create more outliers, where the decision for removing them has to be made cautiously. Beverly Daniels I respectfully disagree. As I posted in the earlier thread regarding control charts: Outlier detection was intended for static data sets (enumerative studies) not for data streams (analytical studies). In SPC a ‘special cause’ is a REAL value – or set of values – that occurs because a condition creates results that are substantially different than that produced by a stable common set of factors that vary in a predictable and ‘controlled’ manner. These results – even when a single value occurs beyond the control limits – are NOT outliers. They are a signal that the process has changed. Regardless of the data set or the type of study, an outlier is a value that truly doesn’t belong in the data set because it is WRONG. Outliers are to be ‘removed’ or censored from the data only when they are validated as: * Impossible results * Misprints or typos * Results from an invalid measurement event * Mis-read measurements Beverly Daniels Outliers that are valid results are real and will occur to us and our Customers and so should not be censored or ignored in any type of study. Outlier detection is a result of comparing results to a theoretical distributional model (that doesn’t actually exist in real life) and the threshold for detecting the so called outlier is too often set at 95% of the distribution. In theory any distributional model will then have 5% of its data that is supposed to lie beyond the 95% threshold. When you have large data sets, you will detect ‘outliers’ that are SUPPOSED to be there per the distributional model. When you have smaller data sets it’s still quite possible to get an ‘outlier’ that is supposed to be there in your data set. Outlier detection using statistical distribution models is a waste of time. If for no other reason than when you detect an outlier you aren’t supposed to do anything with those values unless they meet the criteria listed above for an invalid result. Beverly Daniels We do not remove outliers to achieve a Normal distribution (this is impossible anyway) as we don’t need Normal distributions to analyze the data properly and effectively. Think about the statement “a better analysis”. What does that mean? A clean ideal statistical test? Or the truth about the process we are analyzing? As Shewhart said (and someone here recently quoted) “Probability models do not generate our data, real world processes do.” Our analysis must match the data, we must not twist the data to meet some ideal analysis technique… Sahjahan H May be I can quote some example from IT industry, which can help us to debate more: Server Availability - Due to the server down on a particular day will not mean Server is not performing well overall. This is an Outlier and we need to identify the root cause for the Server down and we need to exclude this outlier and consider the remaining data for the Server Availability SLA calculation. If this is a recurring issue, then we need to consider this outlier into the analysis. Sahjahan H Beverly, you've quoted "Outlier detection using statistical distribution models is a waste of time, as we are not going to do anything with those data unless the mentioned criteria are present" So, what method you will use to deduct the Outliers? Or, you will not consider the term "Outlier" itself? Erik Laufer This is a great question...perhaps, to bring us back to the original post, I would respond with what question are we asking of the data...what is the objective of the analysis/statistics that we are employing? Am I using data for descriptive statistics of what has been? Am I taking data, and projecting results, to others that were not included in the sample? Am I taking the data, and looking to predict what might occur in the future? These are critical to understand...Beverly's distinction in important (experimental/ad hoc analysis vs observational analysis). So, outliers (interesting data) could be assessed via Tukey gates for retrospective/passive data analysis...quasi-EDA/or potentially Tolerance Intervals. Outliers (interesting data) could be assessed via control charts. Outliers (interesting data) could be assessed via Confidence Intervals/Prediction Intervals...it all comes back to the question being posed of the data and how that informs decisions. Regards, Erik Ravi Sharma Before thinking to make a change in process its imperative to look at the outliers also known as special causes and eliminate them. Every process has two factors impacting I.e. Special and common. Common is the inherent variability and requires a lot of efforts infact a change in process. As I mentioned its worth looking at special causes, remove them. Beverly Daniels One of the issues we are experiencing is how we define our statistical words. This isn’t just semantical nit-picking. Our words matter. Outliers are not Special Causes and Special Causes are not Outliers. I know that some people use these words interchangeably but doing so creates confusion, misunderstandings and inhibits learning. These definitions are not my opinion. They come from notable statisticians who selected the names, developed the definitions and the mathematical method by which we determine the existence of that type of result. Shewhart developed the name of special cause for results that came from a system causes that was not the common system performance of a stable process STREAM. Other statisticians developed mathematical tests for the detection of outliers in STATIC data sets. They defined outliers as values that were a substantial distance from the bulk of the data set. Statistical software uses these different definitions. Beverly Daniels The IT example is a good one regarding the mis-use of the term outlier as well as the dangers of censoring outliers that are not invalid results. First, the downtime was the result of an assignable or special cause. The IT group determined it through science, logic and reason – once they determined the cause they knew – or believed – that it was a different causal mechanism than those for the ‘typical’ downtime events. In all likelihood, they didn’t determine that this was a special cause in strict SPC terms through a violation of control limits. (And that’s OK, Shewhart’s intent encompasses both statistical detected special causes and physically detected special causes). The IT Team took the right action to determine the cause and take action to correct and prevent it in the future. However, censoring it from their downtime reporting is WRONG. This is a misuse of ‘outliers’. The event actually happened, it actually effected downtime. It should be in the performance metric. Rune Søndergaard-Hannen Hi Beverly, first of all where do you have the 95% from UCL & LCL are +/- 3 x Stdev => 99,7%. I think we need to discuss the calculation model of the Stdev used for the calculation of control limits too. Is it the AIAG standard or the "normal" statistic calculation. There can be a relative huge difference between those two models and because of that a potential risk for a different number of outliers. I my opinion outliers needs to be explained either by a none normal incident or with a real RCA. Whar are opinion in this forum to this? Christopher Ayres Use correct distributions when posting results. Also make sure outliers are real outliers within your analysis. Make sure they are not cause of bad data collection. If your doing say a perato for a real thing then you must count them in a 0-100 mindset because 1 defect is still a defect however keep in mind if it's mass data 6 sigma still aloes for 3.4 defects per million opportunities. Christopher Ayres Outliers also depend on what results your looking to achieve. In educational econometrics they will just lower your confidence level if even used depending on how far out they are. If your measuring auto defects then an outlier is still a defect, etc. and must be accounted for. Like I said before though 2 outliers out of a 10 thousand study will still give you a confidence level of +or- 99% ate. Ravi Sharma Outliers are the measured observation that don't to seem to fit the grouping of rest of the observation. They are either too far to the right or to the left of the rest of the data for someone to conclude that they come from the same set of circumstances that created all other points. When we see an outlier on a dot plot or histogram we immediately know that something is different about the condition that created those points whether its process set up or execution or the way we measured the process. Investigate all outliers, find out what caused their value to be so different. Rune Søndergaard-Hannen I talk about the confidence interval where 99,7% of all datapoint should be if the process are in control. But yes It's possible to say 95% of the points are inside CL and because of that we have 5% outleirs. That Will just signal that the process aren't in control... If that's the case then a huge job will be in front of you to get the process in control. Investigate potential lines in the reasons etc. to smoothen the process .
  12. The instructor is amazing. Have great sense of knowledge which can be embedded in real time scenarios. - Manjuladevi Krishnan, MSVC, Shelll India Pvt. Ltd.
  13. ahp View File ahp Submitter Karthik Prabhakar Submitted 11/03/2015 Category Admin
  14. AHP View File AHP Test Submitter Karthik Prabhakar Submitted 11/03/2015 Category Public Training Brochures
  • Create New...