FRIDAY, JULY 25, 2014
Font Size
Topic What is the benefit of Rolled Throughput Yield (RTY)

What is the benefit of Rolled Throughput Yield (RTY)

Home Forums Old Forums General What is the benefit of Rolled Throughput Yield (RTY)

This topic contains 9 replies, has 1 voice, and was last updated by Avatar of Steve B Steve B 11 years, 2 months ago.

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #40873

    I’m looking for the hidden factories in a current process as well as defining the defect opportunities. I just finished a process map and I want to know the difference between the documented process and hte actual process. Of course, I’m assuming that some of the workers or managers will not identify the reworks or inspections. I have only data that reports defective (not defects) units through out the process. Is there a benifit with RTY? If so, how do I calculate RTY, from the research that I have done, the calculation and setup is very confusing.

    #40876

    RTY is be very useful.
    Assume 5 step process. 
    Step 1: you have 100 parts.  five are bad, make rework or scrap.  It has yield of 95%, or 0.95.
    Step2: 100 parts output from process, 8 parts defective.  It has a yield of 92% or 0.92.
    Step 3:  Of 100 parts, 4 defective for 96% or 0.96.
    Step 4 : Yield of 90% or 0.90
    Step 5: Yield = 0.99
    Doing the math: 0.95 X 0.92 X 0.96 X 0.90 X 0.99 =  0.75
    Meaning, part has mabe 75% chance of go it through all process steps without defective.  Indivigual yields that some may say “not bad,”  Rolled Throughput Yield get your atencion.  You rework or scrap 25% of everything you builting

    #40899

    RTY simply means the proby of manufacturing a defect free unity. it’s up to you if this metric is important to u and to your company as well.

    #40941

    While I do agree with you on it being another measurement, the RTP concept also comes handy in determining / investigation which of your process step or sub processes are the ones contributing to maximum error rates thus giving you an opportunity for improvement.
    Thanks

    #41028

    I don’t see how the RTP can give you the process or subprocess specific tasks that are contributing to error rates. This RTP calculation appears to be an over all error rate and not a particular process step. If I have to collection information for my 337 process steps for this national project to understand the opportunities, I will never finish this project. Right?
    Doesn’t this calculation look at the overall defect or defective units? Or are you saying that the data that is collected will point back to the process step. My data is not able to point back to a particular step because of the generality of the error description.
     
    While I do agree with you on it being another measurement, the RTP concept also comes handy in determining / investigation which of your process step or sub processes are the ones contributing to maximum error rates thus giving you an opportunity for improvement.

    #41029

    I have some questions with your data collection.
    Step2: Doesn’t Step 1 and 2 over lap in this calcualtion. If the emphasis is on deflective, I would assume that Step 1 is included with bad parts. In addition, I can not segment my data from defects or defective, therefore I can only use defective units..
    Step 3:  Is Step 3 the same as the other steps??
    Step 4 : I still don’t see this calculation, either.
    Step 5: How do you get the 99%?
    I guess I need stat’s 101.
     

    #41030

    RTY calcs aside,
       I would consider following a unit, or representative sample of units through the process from womb to tomb and record, with copious notes, everything that happens to it/them.   In Lean/Agile implementation, I’ve used a hand held video camera, or digital still camera with time counter to capture this information.    Rather than depending on the veracity of the operator/Manager if you suspect less-than-helpful cooperation, find the hidden factory for yourself.
       Basically, “become the unit/part” and walk the line.
     
    My $0.02
    -Heebee

    #41104

    If you have 337 steps in your process, you are way too deep in the details. Start with a SIPOC (4-7 steps). Use RTY to determine which step has the most defects, then break that step into a subprocess of 4-7 steps (or so). Continue to drop down one level at a time until you find your biggest problem and fix that.

    #41178

    RTY – The benefit of it is to show the hidden factory.
    If you have a process with 4 steps and you get at inlet 100 units and at end 85 units you will see the Classical Yield of 85%.
    However you if define Yield as Doing First Time Right suppose you find out in the 4 steps. 
    step 1 has Input 100 units and rework of 7 which results Yield 97%
    step 2 has Input 93 units and scrap 10 which results Yield 89%
    step 3 has Input of 83 units and rework 10 and scrap 3 which results Yield 84%
    step 4 has Input of 70 units and rework 5 and scrap 2 which results Yields of 90%.
    The RTY in this process is = 97%*89%*84%*90= 63% ! quite different from classical Yield of 85%. Classical Yields does not take detect the rework or the “hidden factory”.
    RTY is good to help to focus in which step you need to work on. Find out rework area requires asking questions to the team once the process map is posted in a flip chart. Normaly question is what if?. Another way is to walk in the plant and note for rework stations not included in the process map. 
     

    #41404

    Two points,
    Walking the “part” through the process is the sometimes the only way to identify the hidden factories. Workers/managers who are imbeded in the process are to close to it to be objecttive and often don’t realize they are doing re-work. To them, it is part of the “normal” process.
    As for the calculation, a good example for its use in in IT based projects, since you are often looking at complex transaction based systems. As the transaction passes from one component to another, they perform differently. RTY is the only way to get a true measure of the systems performance. The trick is to measure each component yield, that is the hard part.
     
     

Viewing 10 posts - 1 through 10 (of 10 total)

The forum ‘General’ is closed to new topics and replies.

Login Form