Assembling everything for each item separately

[realm174]

Hmmm Let's see if I can explain this properly... Did someone ever try to have only the basic products on belts, and instead, have an assembler line for each item? Here's what I mean...

Typically, what I've seen is people that have factory lines where they will build certain items that are pushed down the line to be used by multiple other assemblers. For example, one might build green circuits, put those on a belt to be grabbed by any other assembler that needs green circuits.

Would it be more/less efficient to have those green circuits built wherever they are needed instead of being sent down the production line? So basically, for each item that requires a green circuit, have an assembler build one for that particular item?

I'm sortof tempted to start a new game and try that...

[GodricSeer]

I think it would be technically less efficient, since a single belt of advanced circuits is equivalent to many belts of copper/plastic/green circuits. Likewise for other products, they usually get more space efficient as more production has been done. However that shouldn't stop you from trying this. I would love to see a base with an enormous bus of nothing but ore and crude oil, with every intermediate produced locally. You would likely need advanced oil processing first though, unless you bus the different fluids instead of crude.

[userasd]

I think the other way around is more efficient... Have an assembly line for every item and a train/belt transportation to the assembly line that item is needed.

It will require a lot of space and will make your factory a bit complex with all that belts/trains running around so I usually create some basic items on site too (like copper wires, green circuits etc.)

[Frightning]

Isn't that more or less how people typically do Main Bus style bases? I suppose people run a belt or 2 of Steel plates and/or maybe Electronic circuits, but otherwise it's Iron/Copper plates being belted to every recipe block. I personally think Main bus designs are pretty inefficient from a start cost standpoint, and even the fixed costs v. throughput is lacking imo. Building store is a much better way to start off, and then you can go straight to Logistic robots for your large scale base, which allows for a much simpler and better scaling design.

[vanatteveldt]

If you have a set goal, e.g. 1 rocket per minute or X beakers per minute, you can design a custom line leading exactly to that goal.

Sharing assemblers can be more space efficient if not everything is running 100%, but if you are designing a factory to deliver a certain throughput all important parts should be running 100% anyway, so it doesn't matter. The nice thing of making a lot of things in-place is that it is much easier to spot bottlenecks, as otherwise your modules are lacking red circuits, but the shortage can also be caused by a lack of green circuits, etc.

In any case, it's not a black/white choice, as I think no-one centralizes copper cable production so it can be shared by assemblers, but most people centralize smelting. In the extreme "make it where you use it" case you would only transport raw ores and every assembly line would have to smelt (and even make power?) itself.

What I often do is make green circuits separately for most things, but have the red modules make their own green modules, as the ratio is nice (1 green on 8 reds, IIRC) and that way you can immediately see why your red is stalling. You can do the same for blue as it is approximately 1:1:1 green:red:blue (viewtopic.php?f=202&t=42594).

I almost never put gears on the bus, (even though transporting gears is more efficient than transporting plates), but in a lot of cases it is 1:1 gear assembler:X assembler anyway. For producing belts etc it can be more useful to have a separate gears belt as production is intermittent anyway.

[wolletd]

I almost never put gears on the bus, (even though transporting gears is more efficient than transporting plates), but in a lot of cases it is 1:1 gear assembler:X assembler anyway. For producing belts etc it can be more useful to have a separate gears belt as production is intermittent anyway.

In my experience, getting belts sorted from the bus to production lines also is a pain when you got gears there, as most of the time you need plates and gears at the same time. When producing gears on site, you just need more plates from the bus.

[impetus maximus]

for some stuff like gears, and copper wire i find it better to make it locally yes.
i tend to group assemblers together that share ingredients like pipe, or belts for example.
still prefer to bus iron, steel, copper, green/red circuits, processing units, plastic.

[edit] it's a blast to have 'towns' where things are imported, then other things made and exported all via train.

[Tym]

you are describing a "Just-in-time" sort of processing as opposed to the seemingly more common approach of setting up busses of intermediate product. While I have yet to attempt a mega-base of that sort requiring tons of product, there's just -something- about having all that unused product on belts, waiting to be used.

[vanatteveldt]

it's a blast to have 'towns' where things are imported, then other things made and exported all via train.

And one of the big advantages is that with trains you can easily ship as many different products as required without loss of efficiency, while for a bus it makes it much more complex to have 15 goods on the bus and connecting the bus to the assembly lines... (see e.g. the contraptions in dee-'s modular extensible thread viewtopic.php?f=204&t=8377)