# RTY Calculation

Six Sigma – iSixSigma Forums Old Forums General RTY Calculation

Viewing 9 posts - 1 through 9 (of 9 total)
• Author
Posts
• #43197

Al
Participant

Does anybody nave any ideas about how to calculate the Rolled Throughput Yield in cases whre the process is not in series.
I work in a plant where there are several process: stamping (enclosures), which is the feeder process to the Electro coated line; once the enclosures are finiched then they are supplied to the assembly line.
Each process has its own RTY, hoe do you calculate the RTY of the Plant, is that make sense to calculate the RTY for the plant? is that the right metric to keep track of performance?
I would appreciate any inputs.
Regards,
Almudena

0
#136764

Heebeegeebee BB
Participant

Though not in DIRECT series, do these parts integrate down the road?   if so, you can map their insertion points and use RTY calcs.
For the seperate sub-process steps, i’d go with a FPY measurement.

0
#136767

Participant

If your plant only produces one product, then you can calculate your plants RTY by multiplying all of your FPY’s (as a %) that the product goes through, regardless of order.
If your plant produces multiple products though, than a plant RTY won’t really tell you much about your processes.
The key to remember is that your overall process is only as good as its worst/slowest component, a.k.a. bottleneck.  If process A is extremely efficient and has a very high FPY, but process B isn’t as stable and is slower than process A, you are only going to end up with significant inventory due to overprocessing of step A.  Sometimes people get excited about individual efficiency, yet it actually hurts the overall goal.
Hope this helps, best of luck to you!

0
#136787

April
Participant

If make sense RTY for plant, depend by KPIs you defined. How calculate it depend by your plant organization, anyway, RTY for plant can use the same formula of any RTYs, if necessary you can weight different processes.

0
#137090

Jeff
Participant

Al,
Here are my suggestions:
Map out your process.  Determine if parallel paths are AND (both process outputs must be defect free to proceed) or OR (either process outputs must be defect free to proceed) gates.  If AND gates, multiply RTY rates for that portion of the process.  If OR gates, then you add the RTY’s for that piece of the process.
By following this sequence, you can string together each piece of the process and obtain an overall RTY value.

0
#137115

Jered Horn
Participant

Almudena,
I’m guessing the statistical answer is just to have separate RTYs for every (series) path.
So, if you have a process where you stamp and treat/coat component A that gets assembled to component B (that was stamped and treated in a parallel process), then you would have two RTYs…one for component A and one for component B.  That’s assuming component A and component B stamping are truly different processes.  Then you’ll need to look at the coating process(s).  Are the FTYs different for the two components, or is it the same process?
Now, personally, that doesn’t satisfy me.  Like you, I want to have one RTY value for the complete process, whethere there are parallel steps or not.
Here’s what I do…right or wrong.
As an electrical engineer, I know when I have two resistors in series, I simply add the resistence of the two resistors to determine the value of a single equivalent resistor.  So, Req = R1 + R2.  With RTY, we’re multiplying FTYs in series (instead of adding).  So RTY (or FTYeq) = FTY1 * FTY2.  Now, when two resistors are in parallel, we calculate the reciprocal of Req by adding the reciprocals of the resistances of the two resistors.  So, 1/Req = 1/R1 + 1/R2.  So, replacing multiplication with addition, like before, my suggestion is to figure the recriprocal of the RTY of two parallel processes by multiplying the reciprocals of the FTYs…1/RTY = 1/FTY1 * 1/FTY2.
Simple example…
Consider 4 separate process steps (S1, S2, S3, & S4).  S1 and S2 are parallel processes, both feeding into S3.  S4 is a process step in series with S3.  FTYs are as follows: S1 = .80; S2 = .85; S3 = .95, S4 = .90.
“Statistical” Method:
RTY1 = S1 * S3 * S4 = .68
RTY2 = S2 * S3 * S4 = .73
“Electrical Engineer’s” Method:
RTY = (1 / (1/S1 * 1/S2)) * S3 * S4 = .58
Again, that’s what I do.  You probably won’t find it in a stats, Six Sigma, or QC book, but that doesn’t bother me…I got my one number.
Good luck.

0
#160997

Samuel
Member

Almudena,
This is the way I proceed, hope this can help.

Total RTY for Parallel Process

RTYprocessA = FTQstep1   X   FTQstep2  X  FTQstep3   X   FTQstep4    =
0.95

RTYprocessB = FTQstep1   X   FTQstep2  X  FTQstep3   X   FTQstep5    =
0.85

RTYprocessC = FTQstep1   X   FTQstep2  X  FTQstep3   X   FTQstep5    =
0.65

a)  Calculate dpu (defects per unit) for each RTY using    dpu = -LN (RTY) :

dpu processA =
0.051293

dpu processB =
0.162519

dpu processC =
0.430783

b)  Calculate Average  dpu   for the processes:     AVG (dpu processA,B,C)  :

AVG  dpu  =
0.214865

c)  Calculate  RTYTOTAL   using    RTY=EXP(-dpu) :

RTYTOTAL  =
0.80665
=
80,6%

0
#189415

danny
Participant

I can not agree with you more Adam. Why is it that company did so bad then, any ideas. When I took over as Production Manager of manaufacturing and Todd W. was still there; they were only producing 175-225 enclosure a day. Within a month or so manufacturing completed complete enclsoures of 325-350 per shift plus 225 Compaq pershift. Over loded paint line so bad they had to bring on board a weeked paint line shift just to get everything paited. Not hard when you understand manufacturing. I was doing such a bad job they replaced me with someone that had a college degree and guess what happened to production. But I do agree with you Adam with your comments. Well said.

0
#189418

MBBinWI
Participant

Wow, posting a reply to a nearly 4 year old post – and an incoherent one at that.  You have just made a new category all to yourself.

0
Viewing 9 posts - 1 through 9 (of 9 total)

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