I need help on an issue. I created a brewery model that has 2 kinds of beer bottles. If the bottle is green it's taken by a robot and beer is filled by a singleproc, cap is placed by a assembly station and at the end label is placed on the bottle by an assemby station again. The same applies for the white bottle but only difference is that, it is taken by another robot and send to anothler line which has the same processes. Eventually the lines come together at one quality control.
In here my task is to detect 5% of the bottles as defected and send them to trash. In order for me to do that I should create in total of 5% defected bottles in stations filling, cap placing and labeling.
Do you have any proposals how I can do it? Either by a method or any kind of strategy? How can I produce different MU's according to a percentage by using Method at singleproc or assembly stations. Or can I make assembly stations not to place label or cap on the bottles 5% of the time?
I was working on "z_emp" function that I saw in here. By using a method and trying to create 5% defected bottles with different name and catch them with name attribute at the end with quality control. I also added my model if you wanna check.
Thank you very much in advance!
Considering the fact your model obviously doesn't have to care where exactly the scrap is created you could simply follow the approach in the model attached. The 5 % for each kind of bottle will already be assigned to the bottle in the moment it is created at the source and at the end of the line you select and remove all these bottles.
In case the bottles are detected and removed in-line this would of course not work. But to be serious in this case your customer should be able to tell you in more detail which process step causes which amount of scrap. If they can't do so the approach above is the most efficient one.
Thank you for your interest! I was unable to open your model because I am currently using student version which is version 12. Because I am currently a master student and this simulation is a part of my project. So the purpose that I am doing this is educational.
But as far as I understood from your comment, you implemented percentages directly into the source and bottles are gonna arrive the quality control with already 5% defected. If this is the case, it definitely gonna work for my model but let say I was looking for a more realistic option like creating the problem at the station itself. If I have no option doing it, I'll go with your solution.
Thanks again, I appriciated for your help.
I would like the more realistic approach too. But in this case your customer will have to tell you what single scrap rate is created at the stations. Otherwise you only could split the amount of resulting scrap (5 %) freely to the stations. The ExperimentManager could simply run a range from lets say 1 to 3 in steps of 0.25 % for each affected station evaluating the resulting total amount of scrap. There will be several combinations each delivering 5 % in total and you would have to guess which is the best fitting one.