Mayank Gupta

Mohamed Asif is the winner for this question for explaining the concepts of workload balancing and detailing how AI and ML can help us with them. Answer from Sanjay Bhure is also a must read.

Moushmi Kandori has provided the best answer to this question. Mohamed Asif's advise on working with weighted scores is also relevant and is used in many organizations. Many participants have suggested to work with DPMO even with the possibility of working with DPU and Yield%. For a process where there is a possibility of calculating all the three, DPMO will give you the best capability. But does it really reflect the true picture? - Food for thought

While a couple of participants have tried to answer the question, however no one has really gone in details of the specific situation given. Hence, there is no winner for this question. In the situation given here, the team is collecting data on both defectives and defects and defects are also being classified into critical, major and minor. Let's consider an example - we are manufacturing cars. Examples of defectives and types of defects would be something as below Defective car - a car which has defects Critical defect - malfunctioning brakes Major defect - the door does not close properly Minor defect - scratch on the bonnet of the car Stability of the process can be checked with any one of the parameters. From an overview perspective, I would use either a P or an NP chart to check for stability which means I would work with defectives. In addition to this, I would pro-actively work with a C or a U chart for critical defects as there could be multiple critical defects in the same car. By using these 2 charts I will be able to better control my process. Of course, if I have the resources and time, I would like to use C or a U chart even for major and minor defects, however resources and time is a rare luxury

All answers are a must read. Every respondent has highlighted some interesting observation about the Is/Is-Not Analysis. The best answer to this question has been written by Amit Simon - he has provided a relevant example and also highlighted the fact that this tool can also be used to understand a problem.

Pradeep Kandpal has given the best answer to this question. Other answers are also a must read to get varied examples of Transportation and Motion wastes.

All participants have quoted relevant and excellent examples of Overproduction and Overprocessing. The answer that stands out from the others is that of Pradeep Kandpal for providing apt methods to deal with the two kinds of waste.

Kirpa Shanker Tiwari has posted the best answer to the question. While these tests and graphs will help us identify the underlying distribution, there are other methods like goodness of fit tests that can also be used. Most of the statistical packages also have in built functionality for distribution identification.

How much to produce so that neither we over produce nor we under produce is a classical debate in all manufacturing. It is relatively easier wherever demand is constant, but becomes difficult where customer demand fluctuates. In such situations, two approaches have been adopted: 1) demand leveling and 2) production leveling. Both strategies have their own advantages and disadvantages. Depending on the nature of the business and the market conditions, businesses may need to adopt a hybrid approach that balances the trade-offs between these two strategies to achieve best possible results. Best answer to the question has been provided by Pradeep Kandpal. Well done!

Pradeep's answer has been selected as the best answer for the examples that are quoted. Well done!

Pradeep is the winner for this answer. Well done! Do review the answer by Mr Venugopal R, Benchmark Six Sigma's in-house expert

Hello Puneet For sample size determination, we need a little more information 1. What is the purpose of the study? 2. What is the metric under study (e.g. quality, AHT etc) and what difference do we want to detect (e.g. Quality is 90% and we want to detect a difference of at least 5%)? 3. What level of confidence do we need? (we can always work with default of 95%) Regards Mayank Gupta

Hello Gourav Basis the brief provided, I would recommend go with point 1. Objective would be to make the process better as compared to the present condition. SLA determination will happen subsequently. Regards Mayank Gupta

Suresh has explained the most important criteria (volume) along with a few others for deciding the number of bot licenses. His answer has been selected as the best answer.

Balaji has provided the winning answer to this question. Well done!

The winner for this question is Anshul for the clear and detailed explanation of Harada Method. Answer from Nunhuck is an interesting read on origins of this methodology.

Wow. I am truly amazed at the quality of answers that I have got for this tricky and somewhat difficult question. All answers are a must read. There are 2 winners for this question - Anupam Goswami for summarizing the various methods in a tabular format and Nunhuck Oosman for highlighting the drawbacks of randomization.

Very informative answers to an interesting question. Guess everyone's interested in Data Analytics Best answer has been provided by Anupam Goswami.

While there are some good and must read answers (Balaji, Oosman and Keerthi), the best answer has been provided by Suresh Kumar Gupta for also providing some relevant examples of industries were FISH is a problem.

Interesting answers from all participants. Best answer has been provided by Anupam Goswami and Nunhuck Oosman. Answer from Anshul Vaidya is also a must read.

All published answers are correct. The best answer to this question has been written by Vikas Choudhary. Answer from Oosman is also a must read.

Balaji Loganathan has provided the best answer to this question. Response from Oosman is also a must read.

Fleiss’s Kappa and Cohen’s kappa both are used for checking agreements within and between appraisers. While Kappa value can be calculated for any number of appraiser and trial numbers, Cohen's kappa can only be calculated under some specific conditions (e.g. only 2 raters). Also the assumption with Cohen’s kappa is that the appraisers are deliberately chosen and fixed, while with Fleiss’ kappa, the appraisers are chosen at random from a larger pool. The best answer has been provided by Anupam Goswami.

There were many answers to this question. However, not all were published as some of them were incorrect. The question was a tricky one as it dealt with 2 confusing MSA methods - Attribute Agreement Analysis and Attribute Gage Study. Attribute Agreement Analysis is used for discrete data to check for operator agreements on the attributes (there is no gauge being used here). Attribute Gage Study is used when attribute data (like Go/No Go etc.) is being generated using a gauge. The chosen best answer is from Balaji Loganathan. Do review the response from Mr Venugopal R, Benchmark Six Sigma's in-house expert.

There are 2 answers that stand out - Dheeraj Bhardwaj - for explaining the two underlying plots i.e. box plot and kernel density plot Anupam Goswami - for providing insights into where Violin Plot can be used along with examples. Both the answers have been selected as winners. Congratulations to the joint winners!

While all published answers are correct, Anuj Bhatnagar has provided multiple practical reasons for removal of a Red Tag. Hence, his answer has been selected as the winner.