sujeetsingh1

Hi Madan, This is nothing like being a part of group or not, as you understand people spend most of their time in doing their daily jobs, which is of first priority. If i will talk about myself, I login maybe once a week or once a forth night. I will try to explain the DOE with an overview of a typical DOE process. As with any task, the DOE process starts with proper planning. Planning: You need to define purpose of your experiment (e.g., minimize casting porosity). You will need to identify your response variable (response variable = output variable = the "Y" in Y=f(X)), and analyze your measurement system to make sure that your measurement system is capable of detecting differences in your response. Before you start to experiment, you will also have to understand the state of control of your process. Is the process under control? Is it stable? (not necessarily a requirement - sometimes you will experiment with the objective of getting the process back in control) Screening: The purpose of the screening phase is to identify the input variables (input variable = factor = the "X's" in Y=f(X)) that "change" the response variable. First, you will need to find all of your "Potential X's". The next step in screening is to select an experimental design. We need to choose appropriate factor level settings (factor level settings are the values that we will assign to our experimental input variables - e.g., balance spindle rotational speed High/Low = 2400rpm/3600rpm). Finally, we will perform a series of experiments to determine our significant inputs. Finally, we will take our significant inputs into the optimization phase. We will choose an optimization strategy (depending upon our objective, our past history of the process, and the number of significant input variables), and optimize our response. Of course, before making expensive changes or time consuming change-overs, we will validate our findings in a production setting. Enabling Concepts: DoF Strategy Repetitions vs. Replicates Signal to Noise Ratio Interactions Alias Structure / Confounding Hope this will clarify some of your doubts. Wbr, Sujeet Singh

@ Pankaj; I am not able to see the base line target for the process. In my opinion. First of all we need to set a base line in conjunction with software development team for the time process should take to download and upload with respect to the size of the file. Afterwards, compare the current performance with the benchmark goals(SD,mean,sigma level) and find out the gaps. There can be a situation when you may need to set the benchmark on the basis of past best performance with respect to upload and download where you floor ground level executive will be the best source of information as they finish this task on daily basis. Cheers! Sujeet Singh

Dear Jagannath; If this is the situation then, I will recommend you to find out the pain area of your management, even if your organization is not interested in six sigma but they should be interested in over coming there long standing problems. Still if you feel like that, execution of six sigma in ur organization is not possible than you can look around for the firm where you can use your SS skill set and at the same time they have the requirement of people with the same skill set. But still for a fresh GB having one or two completed projects always be advantageous. Good luck!!!

My suggestion would be to stick with the current job and execute one or two projects. There are many reasons to this. for instance... 1. In your current organization bringing changes on the table will not be that difficult as people management will be easy because or prior interaction/support as compare to any new firm with new people. 2. Once you have successfully executed and presented the improvements to the authorities, you might not look around for a new job. Even after executing the improvements successfully, you want to switch over; than you can show case your projects along with your expertise in the related field as explained by Srikanth... Wbr, Sujeet Singh

@ Srikanth S- I have a different point of view on this.... All cost control and cost reduction projects can and can not be directly correlated to customer specifications. At times we have seen customer specifications as secondary metric whereas cost reduction or margin increase can be our primary metric on any given project. With respect to the query raised by Vikram; I agree to Mahendra's remark. Wbr, Sujeet Singh

Dear Anis, Natural process limits are known by several names in the literature: natural process variation, normal process variation, and natural tolerance. In all cases, these terms refer to the +-3σ limits (that is, 6σ spread) around a process average. Such limits include 99.73% of the process variation and are said to be the "voice of the process." Hope this will clarify your doubt on control limits. Secondly your as-is process failure is 5% and you want your process to be targeted at USL 1%, here you can consider LSL as 0% . Your process mean and standard deviation will not change because of change in specification limit. Regards, Sujeet

Hi Sridhar, Your believe is correct, first we need to consider %Tolerance(SV/Toler) if it is acceptable than we need to look for the value of %Study variation(%SV). Regards, Sujeet

Yogesh, If possible conduct 3 or 4 different tests. You will see from your test results that if the T value is below 1.5 or 1 the P value will be high, if your T value is negative your p value will fall under 0.6,0.4 something like this. Which indicates the acceptance of null hypothesis will be high for low T values. Presenting all the tests here will not be possible however I am mentioning T and P values of few tests as mentioned below Test T value P value 1 2.28 0.024 2 19.82 0.00000 3 2.29 0.034 4 -1.17 0.864 One test for the last t & P value. As part of a larger effort to reduce the entire response time when a customer requests special pricing, a new system is being tested to reduce the time to enter a SPA. The new system will be implemented if it reduces the response time by 5.0 minutes. The following data was collected: Manager Proposed Old A 15.5 15.5 B 19.5 15.5 C 23.0 19.0 D 26.5 18.0 E 20.5 15.0 F 17.0 12.5 G 30.5 26.5 H 5.0 5.5 I 7.5 4.0 J 14.5 8.0 Paired t-test and Confidence Interval Paired T for Old - Proposed N Mean StDev SE Mean Old 10 17.95 7.89 2.50 Proposed 10 13.95 6.78 2.14 Difference 10 4.000 2.698 0.853 95% CI for mean difference: (2.070, 5.930) t-test of mean difference = 5 (Vs > 5): T-Value = -1.17 P-Value = 0.864 Hope this will clarify your your doubts. wbr Sujeet Singh

Hi Yogesh, The T value or test statistic for any hypothesis test is actually: (mean - target)/SE mean Often, this value is compared to a critical value. As your T value gets larger, your p-value becomes smaller. Meaning that chances of rejecting null hypothesis increases. Hope this will help! wbr, Sujeet Singh

@Ramesh, This acceptance level is from customer’s as well as inspection stand point; how much tolerance can be given. If you want to go ahead with zero defect with 100% confidence level then you can go ahead with 100% inspection instead of sampling the data for inspection. Sampling method is used to control the inspection cost, else if will be higher. Cheers!

I agree with Harish, the best way as per my experience is to take an initiative in your organization to fix the chronic problem. which is going on for a long. Once you will complete the project by the time you will train your team members on basic SS tools and they will help you in creating the awareness. and also your management will also state taking your inputs positively with respect to SS. Cheers!

Madhan, Normally in a process we consider margin at 5% with 95% confidence level. You have mentioned the statement , which is actually vice a versa. If we will increase the margin then sample size will decrease. The margin of error is the amount of error that you can tolerate. If 90% of respondents answer yes, while 10% answer no, you may be able to tolerate a larger amount of error. Higher confidence level requires larger sample size. The confidence level is the amount of uncertainty you can tolerate. Suppose that you have 20 yes-no questions in your survey. With a confidence level of 95%, you would expect that for one of the questions (1 in 20), the percentage of people who answer yes would be more than the margin of error away from the true answer. The true answer is the percentage you would get if you exhaustively interviewed everyone. In layman language, the process where margin of error can make a huge loss there we keep it very low. Because small margin of error can cause huge losses for example:- one plan crash can cost many lifes etc. Hope this will clarify your doubts. Thanks! Sujeet Singh

Shrikant, Sanjay W has given one example to explain AQL as follows It is generally considered a good plan if for a given criteria of acceptance , say batches with 0.25 % defectives or less are acceptable, then a plan would be expected to be a good one if out of 100 such batches at least 95 are found acceptable by adopting this plan. With the help of AQL, you can decide how many defects can be acceptable with the help of threshold %age.

Hi All, I have a question regarding the calculation of atrition %age. One formula says Attrition=(No of people left/total number of employes on the first day of the month)*100 Second says Attrition=(No of people left/Ave. number of emplyes on first and the last day of the month)*100 Now which is the best and why? Your comments will be highly appriciated! Regards, Sujeet Singh

Process Improvement, Performance Enhacement, Quality Improvement, increased Utilisation of man power. However firstly i will suggest you to find the problem areas in your orgnisation which are effecting the profitability. It can be because of high attrition, low productivity, higher expences on transportation and many more... Regards, Sujeet Singh