Factorio Forums

a few notes:
- on left side of belt are products
- on right side are resources like ores but also plates in this case

- items spread and dilute across the grid like a charm (takes maybe 10-30 seconds to even out on this grid size)
- no micro circulations -> evenly space all different cells and they will be supplied sufficiently

- the belt is a huge buffer !!! which is very good, additional cells and bad adjustments to settings wont crash

- bulletproof? must be tested but experience tells me that holes in the grid are no problem at all

V2 will follow some day

Could you upload a save maybe?
It looks good but its very hard to understand

It's definitly easier than it looks.
If someone has more time than I have got, then please play with it.

But use it properly!
The idea was to cover the whole map with exactly 1 kind of module, obviously that is too easy to be true so I made mines, smelting, assembly and science modules. Later oil or petroleum can fill one of the two pipes.

Here is the safe:

This is awesome, I've adapted it into my current factory design to make miscellaneous products like furnaces, refineries, etc. instead of using robots. Have you done any sort of capacity calculations? In other words, is it possible for this kind of factory design to have a coronary? I've incorporated a "liver" into one of my cells, which pulls everything off of belts and replaces the right amounts.

I'm also working on incorporating a train station into the system directly, with very mediocre results. Once I figure out train station wiring, I'm sure it will be possible to have massive stockpiles that maintain a healthy flow of goods.

Gertibrumm wrote:Later oil or petroleum can fill one of the two pipes.

That is a very fascinating design and I am thinking of using some of your ideas. But i figured out a problem: the oil you mentioned.
You only have one pipe left, but the game has more than one. So how do you want to implement the stuff needing lubricrant, sulfuric acid and so on?

Keeping the current design you could have certain areas supplied with different liquids, AFAIK nothing needs more than 2 liquids to process, but you would need to be extremely careful about keeping them separate. Alternatively with global barrelling coming with 0.15 you could shove it in barrels on the belts, issue then becomes empty barrels clogging your belts but idk how much of a problem that would be.

Alternatively, have all your oil cells in the same place and just underground pipe over the borders. Not really in theme with the design and not exactly expandable but it works.

EDIT: Actually there is a potential problem, not with gaps in the belts but if you completely fill the belts (though this shouldn't really happen) they will stop circulating due to the way splitters work. Basically if both outputs are blocked, it will stop moving, even if one of those outputs loops back around.

All this really means is if you forget to set a maximum or set it too high for something and it floods the bus, it's going to take a little while for everything to start moving smoothly again.

There are some interesting questions about balancing loads too. With research packs on the belts it's probably better to have lab cells spread out for example, even if you just fit them in wherever there's space in other cells, otherwise only one little corner of your factory requires them but it would take a VERY long time for packs at the far end to filter back to the labs.

I played around with your design and I must say, this is very nice for a supply outpost. At the moment, everything except stuff that needs liquid will be produced there. For every ressource I have an extra station and those are put on the outer lane of the belt. The intermediate go on the inner lane. The produced things land in provider chests for the player to fetch at a shuttle station by roboport.

I am working now on another approach to this style. I refuse to let pipes run along with the cells, because it is just annoying to walk there. So I decided to make different regions. Here is stuff that needs lubricant, here goes plastic. This works quiet well as the items distribute quiet equal across the circulation. A problem is, if on one edge you need too much of the same. Therefore you need to restock some items at that place. With basic ressource I had the problem with that restocking, but after I just splitted the belts to inject the iron, copper, steel into different endings of the circulation that worked fine. It is comparable with those 4+ bus system designs.
I also tried out subsystems. I had a science subsystem that took the resources for science packs from the main system and inserted it into the science subsystem. That worked, as expected, quiet good, but I think, that this goes over the top.
Oh, and btw. there were no ores in my circulation. I had my outposts, trains bringing stuff to the furnaces and the plates were thrown into the cell system.

Whatever, thanks for the post, I really love experimenting with that design.

The furnaces get clogged eventually with either stone or iron plates due to those using more than one resource. You need to filter it somehow.

I havnt really had time to test anything but coronary cloggs and bad balance desicions might be the biggest problems but I havnt tested so cannot say for sure.

The two pipes are, as you have guessed, the most essential liquids oil and water. everything on top of that can be produced locally in bigger cells.

ores are there to fullfill all recipes available not just plates

I thought of the whole cells thing more as a proof of concept but it seems to be quite practical (but slow)

I think the furnaces are fine as they only burn copper and iron but not steel, steel would be a completly different type setup.
Maybe connecting the furnaces directly to the net would balance ore and coal a little better

Grimakar can you share your base? I would love to see your developments!

No, the sample map had stone and iron being sent in to create bricks and steel, so it isn't a theoretical problem anyway.

@Gertibrumm Ok, here it is. But keep in mind that there are lots of things under construction. The furnaces need a redesign because I am getting out of space there. Also, the cell system needs to be extended esp for the module part. But I hope you have fun with it and find what you are looking for

Wow thx for the insight!
I really like how you simplified everything!
- no pipes but specialized liquid cells
- no resource/product separation
though I see some cloggs due to unequal distribution on the top side but who cares, its working
- love your science sub module

Gertibrumm wrote:Wow thx for the insight!
I really like how you simplified everything!
- no pipes but specialized liquid cells
- no resource/product separation
though I see some cloggs due to unequal distribution on the top side but who cares, its working
- love your science sub module

Thanks, still there are things I want to change, esp the size of the cells, as seen here: But if I make them 1 tile bigger, then it fits, like on this picture: And to the cloggs you found: That is very strange^^ I am sure there is a simple explanation, why that happens^^

I guess it is best to simply ignore the cloggs.
I think the cell size has to be a compromize:
-Big cells fit more advanced modules but cannot fit small standard productions space efficient
-Small cells fit specific modules like the 8 assembly machine cell but sacfrifice high cost per tile due to lots of belts for nothing

not sure if more crossings aid item distribution

The furnaces clog for a simple reason. Furnaces accept any smeltable item, even if it won't fit. It does not check to make sure it is the same type as the item already there. This is fine for iron and copper ore, but when making steel and stone bricks, you need multiple, so it will drop a single iron plate, then try to put, say, 1 iron ore. The inserter is stuck trying to load ore, and the furnace is waiting for another iron plate.

Also, increasing it by one tile so you can fit a theoretical 9 assemblers inside would make things so much easier.

Yup

As such, if you ignore clogs, the whole furnace setup as is will eventually wind down until you manually clear it. This is clearly a problem. Use filter inserters, or dedicated lines. Or something. Don't know if it is possible to detect what is already in a furnace, but if so, you can abuse that to set a filter. Maybe.

I've built three separate factories with variations of this design, and I have found several things:
1) It is very possible to have a coronary. The easiest way to achieve this is to put ore and smelters on your belt. Red belts have a lot of trouble right around late oil stage, I don't think blue belts would fare much better late game.
2) This is the best thing ever early game. My new build order is power>science>automation>small smelter bank>circuit research>this. Every resource at your fingertips immediately makes many early logistical points moot, and you're still not producing enough to clog the system. And it's pretty cheap, at 160 (?) red and 100 green. It is, however, somewhat expensive up front. I believe the costs are worth it.
3) Remember to get your circuits off the belt as soon as possible.
4) Pipes are unnecessary. Do that elsewhere, over by your smelter/green/red circuit bank.
5) Ideal cell size, in my opinion, is 9x9 (or 11x11, without the corners). This reliably prevents the problem of the first assembler grabbing all the iron by limiting it to two assemblers per "line". Just use different inputs, or complimentary input/outputs.
6) Defense is so easy, just drop a turret next to a belt.

Quick demo of latest (4th) iteration of this design:
https://youtu.be/SJ96_Zw_sx0

Kryptos wrote:I've built three separate factories with variations of this design, and I have found several things:
1) It is very possible to have a coronary. The easiest way to achieve this is to put ore and smelters on your belt. Red belts have a lot of trouble right around late oil stage, I don't think blue belts would fare much better.
2) This is the best thing ever early game. My new build order is power>science>automation>small smelter bank>circuit research>this. Every resource at your fingertips immediately makes many early logistical points moot, and you're still not producing enough to clog the system. And it's pretty cheap, at 160 (?) red and 100 green. It is, however, somewhat expensive up front. I believe the costs are worth it.
3) Remember to get your circuits off the belt as soon as possible.
4) Pipes are unnecessary. Do that elsewhere, over by your smelter/green/red circuit bank.
5) Ideal cell size, in my opinion, is 9x9 (or 11x11, without the corners). This reliably prevents the problem of the first assembler grabbing all the iron by limiting it to two assemblers per "line". Just use different inputs, or complimentary input/outputs.
6) Defense is so easy, just drop a turret next to a belt.

Quick demo of latest (4th) iteration of this design:
https://youtu.be/SJ96_Zw_sx0

Very nice explanation of the system, and thanks for the credit
If I ever find the time, Ill try to iterate once more for an improved version based on some of the issues encountered.