automatic duplicating records for combined printing

[André]

[color:blue]I've got a leftbrain torture for you,

i just hope i can explain te problem clear enough...

Here goes,

I'm using FMP 8.5 Adv for small labels, which are saved as PDF, and in acrobat (with the aid of a imposer plugin) combined on larger sheets, and from there to a printing press.

(All of the above is no problem at all, so you won't find any of this in the samplefile below)

The problem is that each label (item) has its own printrun,

and i'd like to optimise the sheets for the printer as much as i can. Therefore i'm duplicating some records so their printruns are lowered, and with that the surplus of the other items on that sheet are lowered too.

I already made a solution to predict the number of printing sheets, and a calculation for the printrun of each sheet.

What i'm boggling my mind about is a way to automate the process of duplicating records in the most efficient way.

I'd like to note that just duplicating the highest printruns isn't always the best solution, every-time you duplicate a record (what can only be done a limited number of times, because you don't want extra sheets in your press if you don't really need to)

And sometimes it's more efficient to duplicate one of the lower printruns, because of the shifting, the last one of one page would be the first on the next page if a record is added above. And that last one is likely to boost the printrun of that sheet. And it's possible you'll loose there more than you gained above.

As you might have guessed, i've always done the duplication process manually, and it's sometimes a bit of a puzzle which record to duplicate, and keep the number of sheets & printruns optimal.

Now, to save time on future orders, and to make the process easier for others (read: colleagues) i'm searching for a way to automate the optimising process.

I hope all above was clear enough,

suggestions are more than welcome.

I'll include a stripped down sample file

[color:red]Thank you for your help.

combined_printrun_sample.zip

Edited July 21, 2009 by Guest
Sample has now real-world values and its now two files, one not duplicated, and one i've done manually

[André]

[color:red]BUMP

anyone?

[comment]

I don't think your problem is well defined. The goal of the optimization and the constrains are not clear. Is it better to have fewer sheets and more printruns? If so, by how much?

For example, suppose I want to print 3 labels:

"A" x 1500

"B" x 700

"C" x 60

I could give the printers 3 sheets (of 16 labels each), with each sheet containing only one type of label, and ask them to run them 94, 44 and 4 times, respectively, to get:

"A": 16x94 = 1504

"B": 16x44 = 704

"C": 16x4 = 64

for a total of 3 sheets and 142 printruns.

Or, I could make just one sheet, consisting of:

"A" x 10

"B" x 5

"C" x 1

and run it 141 times, to get:

"A": 10x141 = 1410

"B": 5x141 = 705

"C": 1x141 = 141

for a total of 1 sheets and 141 printruns.

And, of course, there are many other possible combinations - but how can one compare them without knowing the cost of each element? And without knowing what can be adjusted (and by how much) and what must remain fixed? (Note that in my example, the quantities produced are only approximately close to the quantities requested.)

BTW, this problem could well turn out to be insolvable or computationally infeasible.