Hi Eric, pass fail sampling plans use individual units of measurement, you can take 315 samples or you can take 316, but it is impossible to sample 315.2 units, because of these rounding errors OC curves will not intersect perfectly except on very unique sampling plans. We could also adjust the reference lines to provide greater resolution on the conclusions but overall the rounding errors are negligible for manufacturing processes. If you have further questions please email us to info@cusum.mx take care!
Hi good morning SED, we did make one six sigma video following DMAIC and using the Minitab tools however it wasn't as popular and we had to take it down.
great video, thank you! AQL vs RQL is a very complicated concept that even senior engineers in my company dont quite get!! Question still remains - how does one determine what is acceptable to the customer? Is this typically an assumption?
Hi Diana, thank you for your question. The Selected AQL/RQL is usually part of a company's risk assessment following corporate guidelines. The tool commonly used is the PFMEA, the PFMEA uses severity, occurrence, and detection values to determine if a defect is high, medium or low risk. Usually, a defect's severity can't be modified without a complete product design change (Costly, time consuming and usually requires regulatory approval). Occurrence of the defect can usually only be changed either with the purchase of better equipment (More capable of consistency) or validating a new process parameter window (With a better performance value Cpk/Ppk) once again it is time consuming and costly. So, when companies find that a defect has a high severity and a considerable occurrence, they use stricter inspection criteria (Detectability) to reduce the likelihood of that defect passing inspection. This usually takes the form of using industry standards such as 0.65 RQL values to assure end user quality and meet PFMEA Criteria requirements. When Defects aren't as high risk (For example 3-4 values in a scale of 1 to 9) lower detectability values tend to be used for example 1.5 RQL. We hope this answers the question, for further feedback please email us to info@cusum.mx We wish you a great day! Sincerely, QE NPI Andres R.
Summary of "Sampling Plans - Minitab Masters Module 2": 1. Module overview: "Sampling Plans - Minitab Masters Module 2" is part of a specialized course for quality and manufacturing engineers, focusing on Minitab usage. 2. Importance of sampling plans: Sampling plans are vital for both quality and manufacturing engineers to ensure product quality and better design manufacturing processes. 3. Key concepts covered: The module covers fundamental statistical concepts such as standard deviation, process mean, and process specifications, along with quality control roles and terminology like AQL, RQL, producer and consumer risk. 4. Box plot analysis: The transcript describes a demonstration of box plots to analyze and compare data spreads, emphasizing the importance of understanding data distribution. 5. Acceptance sampling plans: The module discusses acceptance sampling plans, emphasizing the balance between assuring quality and controlling costs by selecting appropriate sampling sizes and rejection criteria based on defect tolerance.
Hi, this is TG, a quality engineer. Thanks for your video. I have a question for minitab sampling plan vs ISO 2859 standard sampling plan. So our factory usually uses G2 0.1 for AQL standard, could you please explain how this relates to minitab sampling plan? Hope you can see my question and reply. Thanks again.
At 24.0 when you say he will reject with "95%" of confidence if it has 2% defective parts or more. Shouldn't it be 90% rather than 95% or am I misunderstanding?
Hi Ujas, thank you very much for pointing this out. You are correct. The correct statement is: We have 95% confidence that lots with a quality level of 0.65% won't be rejected and we know that lots that are 2% defective or more will be rejected by our client 90% of the time. This 90% comes from our power statement, since we are using a beta of 10 (0.10). If you have further questions please don't hesitate to ask, Thank you once again for your comment Sincerely, QE NPI Andres R.
@@CUSUM-ACADEMY Thank for giving response. 1. I'd love to see any tutorial based on Measurement System Analysis GR&R 2. There is too much confusion about Hypothesis tests and Control charts. One video giving some clarification about which chart can be used and when. Like When can we use t test, X bar R, X bar S, X-MR and such most used charts in Quality Control and manufacturing industry.
@@ujasmodi1890 Hi Please check out Module 4 of our Minitab Masters series on th-cam.com/video/WyORXip9kuo/w-d-xo.html This video includes control charts and common quality tools. Take care! QE NPI Andres R.
@@CUSUM-ACADEMY Yes, I went through it which was on Xbar R and Ppk. Although I am still confused about in which situation which kind charts can be used. And MSA and GR&R are tools where Minitab is used therefore looking for videos on that. I'll be grateful to have that.
@@ujasmodi1890 Hi, we don't have videos on the other bar charts but as a general rule: Use IM-R charts when you take one sample at a time from the production line Use X-Bar Charts when you take multiple samples at the same time (This size is called a subgroup) Use C charts when counting number of defects Use P charts when only two outcomes are possible (Pass/Fail) Take care and keep learning! Sincerely, QE NPI Andres R.
Thanks for the information, very helpful. I have a question, what do you mean when you say 95% of the time? I mean, for example, if we inspect a sample 100 times, we will accept it 95?
Hi Victor! Thank you for the question, it means that if the lot as a whole has less than the AQL level of defects then we will accept it 95% of the time by using this sampling plan, let me give you an example. Let's say you manufacture 100 parts and you have 0 defective parts that means that it doesn't matter how many samples you take because you won't find any defects, now let's say you actually have only 1 bad part in the 100 piece lot, if your sampling plan tells you to take only one part and if it's good accept the whole lot or if its bad reject the whole lot then there is a 99% chance you will accept the lot because it would be very unlikely that that one part you take is actually bad, now lets increase it let's say the 100 piece lot actually has 5 bad parts and you still only take one bad part well in this case there is a 95% chance that you will accept the lot because 95% of the lot is good and you are only taking 1 sample. This is what's known as the AQL level so a 100 piece lot that is 5% defective will be accepted 95% of the time by using a sampling plan of only 1 part to accept or reject. Please let me know if you have follow up questions and I'll do a video on it, take care!
@@CUSUM-ACADEMY Hi CUSUM! Thank you very much for your response and sorry for the late reply. I still don't fully understand, I give you an example: what does it mean to have a batch of 40,000, AQL of 2, Alpha of 0.05 n = 200 and C = 7? According to this, the lot will be accepted 95% of the time if the lot has 4 defective parts but it will be accepted if 7 or less are found according to minitab. This is not clear to me since as you mentioned, 5% would be equivalent to 10 defective parts, it would be very helpful if you could answer. Thank you for your work!
Hi Bassem! Right now the team is really small however we would appreciate your feedback regarding what other videos you would like to see, we are more than happy to help other engineers learn more, be more and earn more!
can we solve this on minitab? can you solve this? Suppose a lot contains 100,000 items. The proportion of defectives p in a good lot is 0.2 with probability 0.8, and 0.4 in a bad lot with probability 0.2. Accepting a bad lot costs $300 and rejecting a good lot costs $200. Sampling costs 1$ per item. Suppose that sampling continues until we observe the kth defective item, at which it stops. Find the optimal value of k
Hello from Costa Rica!! Thank you for the explanation :)
Sorry for the delay, we were on vacation break :) we are always happy to help, if you have additional questions please email us to info@cusum.mx.
very good explanation. thank you to CUSUM .
Fantastic tutorial video here. Very well done. Kepe it up.
Hi great tutorial!
Noticing at 26:22 the reference lines don't precisely intersect with the OC curve. Is there a reason for this?
Hi Eric, pass fail sampling plans use individual units of measurement, you can take 315 samples or you can take 316, but it is impossible to sample 315.2 units, because of these rounding errors OC curves will not intersect perfectly except on very unique sampling plans. We could also adjust the reference lines to provide greater resolution on the conclusions but overall the rounding errors are negligible for manufacturing processes. If you have further questions please email us to info@cusum.mx take care!
Very good instructor...Thank you.
It's our pleasure Diptikant, thank you for your kind comments,
Sincerely, QE NPI Andres R.
Much appreciated! How about a Six Sigma case study using minitab tools, and maybe following the DMAIC approach? That would be cool
Hi good morning SED, we did make one six sigma video following DMAIC and using the Minitab tools however it wasn't as popular and we had to take it down.
@@CUSUM-ACADEMY That's a bummer, I wonder if you could reupload it, it would very helpful to me and for sure to many others :)
great video, thank you! AQL vs RQL is a very complicated concept that even senior engineers in my company dont quite get!! Question still remains - how does one determine what is acceptable to the customer? Is this typically an assumption?
Hi Diana, thank you for your question.
The Selected AQL/RQL is usually part of a company's risk assessment following corporate guidelines.
The tool commonly used is the PFMEA, the PFMEA uses severity, occurrence, and detection values to determine if a defect is high, medium or low risk. Usually, a defect's severity can't be modified without a complete product design change (Costly, time consuming and usually requires regulatory approval).
Occurrence of the defect can usually only be changed either with the purchase of better equipment (More capable of consistency) or validating a new process parameter window (With a better performance value Cpk/Ppk) once again it is time consuming and costly.
So, when companies find that a defect has a high severity and a considerable occurrence, they use stricter inspection criteria (Detectability) to reduce the likelihood of that defect passing inspection. This usually takes the form of using industry standards such as 0.65 RQL values to assure end user quality and meet PFMEA Criteria requirements.
When Defects aren't as high risk (For example 3-4 values in a scale of 1 to 9) lower detectability values tend to be used for example 1.5 RQL.
We hope this answers the question, for further feedback please email us to info@cusum.mx
We wish you a great day! Sincerely,
QE NPI Andres R.
Summary of "Sampling Plans - Minitab Masters Module 2":
1. Module overview: "Sampling Plans - Minitab Masters Module 2" is part of a specialized course for quality and manufacturing engineers, focusing on Minitab usage.
2. Importance of sampling plans: Sampling plans are vital for both quality and manufacturing engineers to ensure product quality and better design manufacturing processes.
3. Key concepts covered: The module covers fundamental statistical concepts such as standard deviation, process mean, and process specifications, along with quality control roles and terminology like AQL, RQL, producer and consumer risk.
4. Box plot analysis: The transcript describes a demonstration of box plots to analyze and compare data spreads, emphasizing the importance of understanding data distribution.
5. Acceptance sampling plans: The module discusses acceptance sampling plans, emphasizing the balance between assuring quality and controlling costs by selecting appropriate sampling sizes and rejection criteria based on defect tolerance.
Hi, this is TG, a quality engineer. Thanks for your video. I have a question for minitab sampling plan vs ISO 2859 standard sampling plan. So our factory usually uses G2 0.1 for AQL standard, could you please explain how this relates to minitab sampling plan?
Hope you can see my question and reply.
Thanks again.
At 24.0 when you say he will reject with "95%" of confidence if it has 2% defective parts or more. Shouldn't it be 90% rather than 95% or am I misunderstanding?
Hi Ujas, thank you very much for pointing this out. You are correct. The correct statement is: We have 95% confidence that lots with a quality level of 0.65% won't be rejected and we know that lots that are 2% defective or more will be rejected by our client 90% of the time. This 90% comes from our power statement, since we are using a beta of 10 (0.10). If you have further questions please don't hesitate to ask,
Thank you once again for your comment
Sincerely, QE NPI Andres R.
@@CUSUM-ACADEMY Thank for giving response.
1. I'd love to see any tutorial based on Measurement System Analysis GR&R 2. There is too much confusion about Hypothesis tests and Control charts. One video giving some clarification about which chart can be used and when. Like When can we use t test, X bar R, X bar S, X-MR and such most used charts in Quality Control and manufacturing industry.
@@ujasmodi1890 Hi Please check out Module 4 of our Minitab Masters series on th-cam.com/video/WyORXip9kuo/w-d-xo.html
This video includes control charts and common quality tools.
Take care!
QE NPI Andres R.
@@CUSUM-ACADEMY Yes, I went through it which was on Xbar R and Ppk. Although I am still confused about in which situation which kind charts can be used. And MSA and GR&R are tools where Minitab is used therefore looking for videos on that. I'll be grateful to have that.
@@ujasmodi1890 Hi, we don't have videos on the other bar charts but as a general rule:
Use IM-R charts when you take one sample at a time from the production line
Use X-Bar Charts when you take multiple samples at the same time (This size is called a subgroup)
Use C charts when counting number of defects
Use P charts when only two outcomes are possible (Pass/Fail)
Take care and keep learning!
Sincerely,
QE NPI Andres R.
Thanks for the information, very helpful. I have a question, what do you mean when you say 95% of the time? I mean, for example, if we inspect a sample 100 times, we will accept it 95?
Hi Victor! Thank you for the question, it means that if the lot as a whole has less than the AQL level of defects then we will accept it 95% of the time by using this sampling plan, let me give you an example. Let's say you manufacture 100 parts and you have 0 defective parts that means that it doesn't matter how many samples you take because you won't find any defects, now let's say you actually have only 1 bad part in the 100 piece lot, if your sampling plan tells you to take only one part and if it's good accept the whole lot or if its bad reject the whole lot then there is a 99% chance you will accept the lot because it would be very unlikely that that one part you take is actually bad, now lets increase it let's say the 100 piece lot actually has 5 bad parts and you still only take one bad part well in this case there is a 95% chance that you will accept the lot because 95% of the lot is good and you are only taking 1 sample. This is what's known as the AQL level so a 100 piece lot that is 5% defective will be accepted 95% of the time by using a sampling plan of only 1 part to accept or reject. Please let me know if you have follow up questions and I'll do a video on it, take care!
@@CUSUM-ACADEMY Hi CUSUM! Thank you very much for your response and sorry for the late reply. I still don't fully understand, I give you an example: what does it mean to have a batch of 40,000, AQL of 2, Alpha of 0.05 n = 200 and C = 7?
According to this, the lot will be accepted 95% of the time if the lot has 4 defective parts but it will be accepted if 7 or less are found according to minitab. This is not clear to me since as you mentioned, 5% would be equivalent to 10 defective parts, it would be very helpful if you could answer. Thank you for your work!
hi my name is bassem.
i am a quality control engineer i would like to join your team
Hi Bassem! Right now the team is really small however we would appreciate your feedback regarding what other videos you would like to see, we are more than happy to help other engineers learn more, be more and earn more!
can we solve this on minitab?
can you solve this?
Suppose a lot contains 100,000 items. The proportion of defectives p in a good lot is 0.2 with probability 0.8, and 0.4 in a bad lot with probability 0.2. Accepting a bad lot costs $300 and rejecting a good lot costs $200. Sampling costs 1$ per item. Suppose that sampling continues until we observe the kth defective item, at which it stops. Find the optimal value of k