Comments on: Calculating Rolled Throughput Yield (RTY)

By: Santiago

Santiago — Mon, 08 Feb 2010 22:19:15 +0000

Yes, my question its more focused on the actual collection of the data, since it has several processes, i need to have one person on each process just collecting the data? since the actual operator will be busy with his operation.

Thanks!

By: Ron Pereira

Ron Pereira — Mon, 08 Feb 2010 22:13:06 +0000

Hi Santiago, assuming you have different processes within the cell you would simply calculate the yield at each process and then multiply them together as described in the article above. Make sense?

By: Santiago

Santiago — Mon, 08 Feb 2010 22:09:39 +0000

How do i establish a method to measure RTY in a cell?

By: Top 10 LSS Academy Articles in 2009 (Pageviews) | Lean Six Sigma Academy

Top 10 LSS Academy Articles in 2009 (Pageviews) | Lean Six Sigma Academy — Fri, 18 Dec 2009 12:56:14 +0000

[...] Calculating Rolled Throughput Yield (RTY) [...]

By: Bill

Bill — Mon, 02 Nov 2009 16:14:17 +0000

is RTY designed to measure the odds of just a piece part passing during a sequence of operations or can it be used to meaure the odds of an assembly and it’s related sub assemblies passing? IE if I have a production cell that builds 5 sub assemblies that comprise 1 compelte unit, can I use RTY to tell me how my cell is performing? The examples I have seen always talk about a part going from one process to the next and parts either being reworked or scrapped or passing. That is simple to understand, my example is more complex as there are hundreds of parts in the completed unit. Would you recomend that I substitute each of my sub assemblies for a “process” and multiply the passing % of each subassy times each other to get RTY? How would I include tests on the completed unit? thanks.

By: Lusi

Lusi — Thu, 22 Oct 2009 18:05:24 +0000

Hello All,

could you tell me how can i improve RTY if most of defects are caused by defectived parts (provided from supplier, for example parst which lead to leakage) or functional defects ?

By: Ron Pereira

Ron Pereira — Fri, 04 Sep 2009 22:03:40 +0000

Good question, Sergio. I’d probably split them. So we’d have a RTY for parallel process 1 + series system and then a RTY for parallel process 2 + series system. Make sense?

By: Sergio Lopez

Sergio Lopez — Fri, 04 Sep 2009 21:52:44 +0000

Whats if the processes are in parallel and then feed into a series system?

By: Ruddy

Ruddy — Wed, 03 Jun 2009 16:33:06 +0000

RTY – Rolled throughput yield as I can see is more concerned with how effective the current process is. It accounts for how many parts are accepted with or with out defect. Unacceptable parts are scrapped. Therefore this is a measure of yield throughout an entire process which is a different measure than DPU and DPMO.

By: Kent V

Kent V — Tue, 07 Apr 2009 03:32:19 +0000

Isn’t what Mark is getting at is that these methods tend to blur the difference between yields based on overall defects per unit rather than numbers of units that manage to pass at various process steps? Whenever the multiplied yields method is displayed for RTY, people have a tendency to assume each entry represents the yield of good units rather than the long term probabilistic absence of defects among units at each step. There is a subtle but important difference here that demands that we get more specific about “yield” — of what, relattive to what?. Mikel Harry has a cogent explanation of this in his Six Sigma Breakthrough Management book (2000, pp. 83-88.) I haven’t found a better explanation anywhere. I’m still looking.
To flip this around the other way, consider a unit with 10 potential defects. In one sample of 10 units there may be 5 defects all stuffed into one unit. The good “unit” yield would be 0.9. In another sample of 10 with 5 defects there may be one defect on each of 5 units, giving a good “unit” yield of 0.5. But both samples have the same probabilistic level of defects. The reference to yields on a unit basis is of little use in these cases and even shrouds the truth about the extent and risk of defects. This goes way past the confusion about which WIP belongs to which sample batch and how a step-after-step unit yield might be hard to track– a legitimate challenge in any real world plant or heavy transaction environment.