Q 554. Suppose we have a scenario where data is being gathered for both defects and defectives in a specific process. In addition, the defects are categorized into minor, major, and critical, which are the three types of defects that various organizations typically recognize. In this situation, how would you check for the capability of the process? Which capability metric would you prefer to use - DPMO, DPU or Yield%? Please provide examples to support your answer.

 

Note for website visitors -

• This platform hosts two questions per week, one on Tuesday and the other on Friday. 
• All previous questions can be found here: https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/. 
• To participate in the current question, please visit the forum homepage at https://www.benchmarksixsigma.com/forum/.
• The question will be open until the following Tuesday or Friday evening at 5 PM Indian Standard Time, depending on the launch day. 
• Responses will not be visible until they are reviewed, and only non-plagiarised answers with less than 5-10% plagiarism will be approved. 
• If you are unsure about plagiarism, please check your answer using a plagiarism checker tool such as https://smallseotools.com/plagiarism-checker/ before submitting. 
• All correct answers shall be published, and the top-rated answer will be displayed first. The author will receive an honourable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term.

DPMO is the best metric to use, though we can also check the yield %.  Below is one of the project example that I initiated.

 

 

 

 

DPU is measure of average number of defects per unit.

DPMO is the measure of number of defects per million opportunities.

Yield % is the measure of proportion of products that pass the assessment or QC stage.

 

DPMO or DPU are used when the defects are discrete in nature and when it can be counted on a unit basis. Yield% is used when defects are assessed on a pass/failure, true/false basis, mostly in quality control or during quality inspection.

 

Appropriate metric depends on the specific context, specific goals and the requirements of the processes in the organization and can differ from case to case basis.

 

Minor, major and critical defects may be weighted differently in terms of their impact on the overall quality of the prod or service. So, it is recommended to use weighted metrics that accounts the severity of each defect.

 

Software development example:

When we are tracking number of defects in the code and when we identify 100 defects out of 10000 lines of code. DPMO would be:

DPMO = (100 / (10,000 * 1,000)) * 1,000,000 = 10,000

 

For DPU, when we have identified 50 defects in the user interface and when the mobile application has been downloaded and installed on 1000 devices. The DPU would be:

DPU = 50 / 1,000 = 0.05, meaning on average there are 0.05 defects per mobile device.

 

Yield %, during QC, when we run tests on 100 instances with 90 passing the test. The Yield% would be: Yield % = (90 / 100) * 100 = 90%

 

To conclude, when we track both defects and defectives, we can use DPU or DPMO to measure the number of defects per unit or per opportunity and over an above use yield% to measure the proportion of defect free units.

So, choice of metric depends on:

• Specific context / goals
• Type of defect and
• Impact on overall quality

Quote

"The ultimate goal is Zero defects." - Philip B. Crosby

 

DPMO method is useful when we have defects count, samples count, OFE (Opportunities For Error) details. DPMO = total defects /total OFE X 10^6. Total OFE is computed as  number of samples X OFE. DPU method will be used when we do not know OFE. Number of defects is known, OFE not known, number of samples is known then DPU method will be used.

Example: 500 mobile phones inspected, every scratch is a defect, total of 8 scratches found on 5 phones; remaining 495 phones no scratch. defects = 8; sample = 500; DPMO = DPU X 10^6.

Yield % will be used in case if Defectives are known, number of defects or number of sample not known, OFE not known, then DPMO = defective X 10^6. Example: samples of among 1500 bottles, 125 bottles are bad, then yield = good parts / total parts; yield = % good parts/total parts; FTR yield = First time right parts / Total parts.

On 4/7/2023 at 6:25 PM, Vishwadeep Khatri said:

Q 554. Suppose we have a scenario where data is being gathered for both defects and defectives in a specific process. In addition, the defects are categorized into minor, major, and critical, which are the three types of defects that various organizations typically recognize. In this situation, how would you check for the capability of the process? Which capability metric would you prefer to use - DPMO, DPU or Yield%? Please provide examples to support your answer.

 

Let us try and Defects vs. Defectives

A Defective deliverable can have multiple defects and a defect is qualified as inability to meeting the client requirements.

 

Yield: It helps us to calculate the percentage of non defective items produced that meet the client requirements.

 

Yield can be used to assess process capability for the processes that are stable i.e., deliver consistent output as the process operates within a defined range of values.

 - A 6 sigma process produces 99.9997% of the times defect free / non defective  products that meet client requirements.

 - A 5 sigma process produces  99.9770% of the times defect free products

 - A 4 sigma process produces 99.3790% of the times defect free products

 - A 3 sigma process produces 93.320% of the times defect free products

Example - Creating a PBI dashboard. Each screen / view is similar in nature i.e., number of charts, filters, UI etc. of the Dashboard. Each screen there are 200 opportunities for a defect and there are 5 screens to be developed. There are a total of 20 defects found. Then Yield is calculated as 

Total Opportunities = 200 * 5 = 1000

Total #defects = 20

Yield = ((1000-20)/1000) *100 = 98% and we can say process is operating at 3.55 sigma level.

 

DPMO - It provides the rate at which defects can occur in a product / service

Example - Efficiency of the Internal Dashboard Validation process

Validation Checklist is created and used by project teams to validate the dashboard prior to client review. The checklist has 20 checkpoints. The team selects 5 use cases to validate and find a total of 10 defects. We want to analyze the process capability of the internal validation process i.e. number of defects per use case. 

DPMO is calculated as (10 / (5*20)) * 10,00,000 = 100,000 defects per million use cases or 1 defect every 10 use cases and the process is operating at 2.7 sigma level.

 

DPU - It provides the average # of defects per unit produced.

 

Example - In a software application code if 300 lines of code are written with 600 defects then DPU = 600/300 i.e. 2. 

DPU and DPMO are the preferred metrics to be used in the scenario provided in the question 

 

On 4/7/2023 at 6:25 PM, Vishwadeep Khatri said:

 

 

 

I think in this scenario DPMO would serve as a better capacity metric.  DPU is defects per unit however DPMO gives a value of defects per million opportunities. Comparing defects and defectives, where the latter is where the entire unit is not acceptable, calculations on defects are more precise. Dpmo provides a mesure on capability of the process to generate defects which can be used to calculate sigma level. So in case of a discrete data dpmo will be a better process capability metric. Eg soft drink bottles considering the cap, label and bottle diameter, the capability of the process can be better measured using dpmo

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