Longer cycle time

All topics on coding 4Dscript in Enterprise Dynamics.
WUR
Posts: 21
Joined: Monday 10 June, 2013 - 14:14

Longer cycle time

Post by WUR »

Hi all,

I need your help!
I have an assembler with cycle time:

Code: Select all

Max(0, Normal(mins(Parameter(7,1)), mins(Parameter(17,1))))
Whereas 7,1 = 240 and 17,1 = sqrt (240)
The process should be done at 4pm and when I select run until stop time 16.1h (starting at 0.00), the process is done.

Then I have a queue, a server (cycle time = 0) and a queue, followed by another server with cycle time:

Code: Select all

Max(0, Normal(mins(Parameter(16,1)), mins(Parameter(19,1))))

Whereas 16,1 = 180 and 19,1 = sqrt (180)

However, this process end not at the expected 7pm, but at 10pm (22.1h).
How is this possible? The Cycle time of the server in between is 0, there are only some labels added and tested.

Kind regards,
WU, R.
menno
Posts: 45
Joined: Tuesday 29 March, 2011 - 16:01

Re: Longer cycle time

Post by menno »

Hello WUR,

I tried your pieces of code and it looks good, so I think the fault must be somewhere else.

The total cycle times sum up to ca. 7 hrs, is that correct?

I do not understand why you say that the assemble process takes until 16 hr (I think it should be 4 hr)

Maybe you can give some more info on when your assembler starts working (not at time 0 I guess).

Kind regards,

Menno
WUR
Posts: 21
Joined: Monday 10 June, 2013 - 14:14

Re: Longer cycle time

Post by WUR »

I add my model to this post.
I tried to copy a small part, but order sizes etc. are dependent of other atoms.

My assembler starts at 12h when the first order arrives
This takes 4 hours, which is right when you check it at 16h
Then transport should be done at 19h, because mean cycle time is three hours, but this process is done at 22h instead of 19h.

I already changed the parameters in seconds, but that is also not the mistake...
I hope you can find the bottleneck!

Many thanks in advance!
Attachments
Trial 15-07.mod
(122.16 KiB) Downloaded 319 times
menno
Posts: 45
Joined: Tuesday 29 March, 2011 - 16:01

Re: Longer cycle time

Post by menno »

Hello WUR,

I just tried to stop the run after the first product leaves server 'Transport to DC' (the one after Queue 58). This happens at at exactly 19 hours. So to me, it seems to be working just fine. Maybe I do not understand your question.

Kind regards,
Menno
WUR
Posts: 21
Joined: Monday 10 June, 2013 - 14:14

Re: Longer cycle time

Post by WUR »

Hi Menno,

Thanks for your quick respond!
When I run the model the last product enters the server 'Transport to DC' at 19h (at 19.1h all products are in the server), and the first product leaves the server 'Transport to DC' (so enter the queue storage DC) at 22h. While this should be 19h.

I've outlined the time schedule, and then you will see there is a gap of three hours:
8-12h production
12h order received
12-16h pick order
19-22h tranport to DC

While this transport should take place from 16-19h, when you sum up the cycle times to the opening time of the producer.

Regards,
menno
Posts: 45
Joined: Tuesday 29 March, 2011 - 16:01

Re: Longer cycle time

Post by menno »

Hello WUR,

As 'Transport to DC' server takes +/- 3 hours, it takes 3 hours after 16 hr for the first product to be ready at this server. After this the next product can enter etc.

I noticed the following on the batch characteristics of your server:
1+content(in(1,c))

This means, ideally, that the entire batch that is in the queue +1 should enter the server. This does not happen the first time!
Therefore the 'gap' in time difference exists.

The first time only 1 product enters the 'Transport to DC' server instead of the 11 product that you want. This happens because the 'batch' attribute is fired when the queue is still empty (due to the order in the event list). So you should handle this differently.

You can put the batchsize on a label of the first product of a batch, in this way it should work!

Kind regards,
Menno
WUR
Posts: 21
Joined: Monday 10 June, 2013 - 14:14

Re: Longer cycle time

Post by WUR »

Thanks Menno for your explanation!

It is clear to me! However, I'm trying to put a label on the first product of a batch, but I am wondering how I should manage this?

Regards,
Wu
menno
Posts: 45
Joined: Tuesday 29 March, 2011 - 16:01

Re: Longer cycle time

Post by menno »

Try putting:

Code: Select all

do(  
  if(
    content(c) = 2,
    label([batchsize], rank(2,c)) := 11
  )
)
on the trigger on entry chn. 2-n of the assembler and put:

Code: Select all

label([batchsize], first(c))
on Batch(B) van the server.

This should do the trick!

Kind regards,
Menno
WUR
Posts: 21
Joined: Monday 10 June, 2013 - 14:14

Re: Longer cycle time

Post by WUR »

Thank you very much!

I just have one comment. Because now, Batch size is 11, however this should be the same as the order size, because it is varying per day. Now, there are 10 products ordered and 11 delivered.

I tried to change 11 in the formula into 'label([order size DC])', but that did not work...
menno
Posts: 45
Joined: Tuesday 29 March, 2011 - 16:01

Re: Longer cycle time

Post by menno »

Hello,

You should be able to make batch size dependent on the order size (can be done with a label as you suggested). But you must know where to read this information from. I do not know where you determine the order size? On the assembler?

kind regards,
Menno
Post Reply