Reliability Growth: Concepts, Strategy, Duane Model and Application Case Study

We are happy to release this video on Reliability Growth which is a very important strategy to assure reliability of new products. In this video, Hemant Urdhwareshe has explained the basic concepts and strategy of reliability growth, the related mathematical relationships, and the Duane Model with an application case study of introduction of new model of Diesel engine. Hemant is a Fellow of ASQ and is ASQ CRE, CMBB, CSSBB, CQE and CMQ/OE. We are sure that viewers will find it useful. If you like the video, do not forget to click the like button. If you want to watch more of our videos, please subscribe to our channel and click on the bell icon.

[@procubing8071]

All the heros are not wearing cap.. Sir indeed you r a great teacher.. Such a clear presentation..

[@instituteofqualityandrelia7902]

Thank you for your appreciation!

[@mahendralokhande34]

Thank you so much for your videos. They clearly explain the complex reliability concepts in a short concise manner

[@instituteofqualityandrelia7902]

Hello Mahendra, you're very welcome! Appreciate your feedback!

[@sadanandrenapurkar4897]

Thank you. Very crisp explanation!

[@instituteofqualityandrelia7902]

I am glad you found it useful! Thanks!

[@manofhat676]

thanks, now i understand how reliability growth tests work

[@instituteofqualityandrelia7902]

I am glad to know! Thx.

[@pedrohuacause2706]

Interesting video...!!!! Thanks very much....

[@instituteofqualityandrelia7902]

Glad you liked it!

[@cigam3391]

Thanks a Lot, i am from Perú and i enjoy your vídeos!

[@instituteofqualityandrelia7902]

Glad you like them!

[@luciandersonmarquesferreir1165]

Excellent explanation, Thanks!

[@instituteofqualityandrelia7902]

I am glad it was helpful!

[@HARPREETSINGH-nh4fv]

thanks sir for sharing sir it is very useful .

[@instituteofqualityandrelia7902]

Appreciate your feedback! 🙏

[@flutelearner8449]

very clear explanation sir.. thanks

[@instituteofqualityandrelia7902]

Thanks and welcome!

[@anilbp]

Very informative sir

[@instituteofqualityandrelia7902]

Thanks Anil Prasad!

[@subratprajapati1]

Thanks, there are more better practical way to teach why we use these equation rather simply follow whats been written.keep it up

[@instituteofqualityandrelia7902]

Hope you like our videos! Thx.

[@AceHardy]

👑

[@instituteofqualityandrelia7902]

Thanks!🙏

[@sathyafit1407]

Nice example

[@instituteofqualityandrelia7902]

Thanks! Appreciate your feedback!

[@DonMeaker]

The tendency to be a straight line may also be a result of the change in scale as one moves right across the graph. J.T. Duane initially plotted failures vs. time, but was asked "If reliability is improving, why does the line go down?" Next month he plotted MTBF rather than failure rates.

[@instituteofqualityandrelia7902]

Hello Don,
Greetings! My apologies for late response.
Appreciate your comment and information.
With best wishes..Hemant

[@harryzhang4239]

Hi Sir, what if we're studying a system, for example, computer? There're so many components in it. Do we need separate the data by diff. failure modes or just combine them together? Thanks!

[@instituteofqualityandrelia7902]

You need to separate the failure modes for taking improvement actions. For growth calculations, all failure modes should be considered.

[@harryzhang4239]

@@instituteofqualityandrelia7902 Thank you so much for the quick answer. I'm studying the warranty. So I need consider all failure modes. Rght?

[@samirbenammar3177]

Hi,
Are there other tests (such as ALT) when the growth reliability test reaches an acceptable value (expected MTBF)? In other words, is the diesel engine subjected to accelerated testing after the growth reliability test?

[@uhemant1]

ALT is done on subsystem or component level. It is usually done before the growth test which is done on a complete system usually.

[@samirbenammar3177]

@@uhemant1 Thank you for your response

[@prajotnagapure5134]

Hi Sir great Video. Can we use this model on existing machine line in a plant.

[@instituteofqualityandrelia7902]

Reliability Growth technique is generally used for new product development. It may not be useful for existing machines. The growth is achieved by design improvements which may not happen in existing machine.

[@jacobthangaraj3782]

Great video. How to consider failure fix in the calculation? Can we consider this model for reliability demonstration test where the failure got fixed during the test? Kindly suggest

[@instituteofqualityandrelia7902]

A failure that is fixed on ALL systems under test will speed up the growth anyway. In reliability demonstration, Growth model is not appropriate as your system is not being redesigned but only being demonstrated for a given reliability.

[@DonMeaker]

@@instituteofqualityandrelia7902 For estimation of failure fixes, one can use a 'fix effectiveness factor'. If reasonable calculations show the failure mode will occur half as often, the fix effectiveness factor is 0.5, and you count half a failure as remaining, and this boosts your prediction of the 'inherent' reliability. In one case, we implemented two corrective actions: one reduced the frequency, the other was a maintenance policy shift that permitted repair at Direct Support level rather than General Support, and the count of failures that would drive evacuation to General Support was completely eliminated. The reduction in failure rate was necessary to justify the change to the maintenance policy.

[@instituteofqualityandrelia7902]

@@DonMeaker Thank you for sharing your experience. This will be useful to the users. Appreciate your keen interest and time! Regards..Hemant

[@harihattak]

Thanks for thr video...I have a doubt.. Both X and Y axis have been converted to ln(natural logarithm) in tables...but in sample calculation both t and mtbf is being converted to logarithm(base of 10)...Can u pls clarify on this?

[@uhemant1]

Good question. You can use ln as well. I think the video should have been with ln only. However, the constant changes. The slope alpha which is the growth rate remains same.

[@harihattak]

@Hemant's Accordion Delight , Yes sir...I tried with both ln and log...There is no difference in slope..but ln(b)- Y intercept at t=1, was observed to be varying much...

[@mohamedsaber81]

Great vedio,and I have one question ,in the example ,there are 2 failures in second month,it means one failure in first month and one failure in second month ?or both failure in second month?thank you

[@instituteofqualityandrelia7902]

The table shows cumulative failures. So it is 1 failure in first month and one failure in second month. Thanks for your interest!
Hemant

[@mohamedsaber81]

Thank you

[@cristianoiasi7514]

Hi, this is a great video. However I have one question. What happens if the slope goes down. Ie top left to bottom right direction I would assume this is shrinking.

[@instituteofqualityandrelia7902]

Yes! If the reliability is going down (shrinking), the slope will be negative or downwards.

[@cristianoiasi7514]

@@instituteofqualityandrelia7902 thanks 🙏

[@Amit-ki-Pathshala]

I have few questions:
1.Is Software Reliability and Reliability growth analysis are similar?
2. During development stage, all the features of the product doesn't implemented so there are more chances of bugs when testing continues. It may gives us negative growth. So when should we collect development data for growth analysis?
3. Shall we get any MTBF numbers after software Reliability to plug them into end to end Reliability model?

[@instituteofqualityandrelia7902]

Thanks. Apologise for delayed response. I have not worked on software Reliability but it may apply! Try to apply to a software project data.

[@Amit-ki-Pathshala]

I am happy to see now a lot of people are taking interest on this informative video. I am very grateful to you Sir.

[@Amit-ki-Pathshala]

I am eagerly waiting to watch similar video on crow amsaa Reliability growth model.