Factorio Forums

This is a design I have been working on for the past couple weeks. I call it Sushi wagon. I have not seen similar designs elsewhere, at least for me it is completely original.
The basic idea is to move most high-volume items by trains, but allows mixed item types in each wagon:

Items are fetched directly from wagon to assembler, then back to the same wagon. This reduces the intermediate transportation required, and in turn improved efficiency.
The end products are then moved to destination research center, which is a mix of bot and belt design.

Trains have identical schedule, and move a mixture of items to the destination.

As most items are pre-processed, only very few bots are required to operate:
It is quite UPS efficient. I have cloned the module 40 times to reach 12K+ spm @ 60 UPS with still pretty good margin.

It appears to me to are deleting some items from the train In the third to last photo. Why is that?

i have trouble buying that this works at all, considering i've tried it with Bob's mods and train-to-train-to-train builds where I use beacons and productivity modules and every time things go out of balance because you can't ensure each assembler has the same level of productivity boost, as different number of overlapping beacon ranges affect each.

can you at least show the train schedule? you mentioned it, but didn't give any hints as to what it looks like.

ah, nevermind. you're using bots. that is an impure setup. it will not go out of balance because the bots don't let it.

NotRexButCaesar wrote: ↑

Sun Mar 21, 2021 10:25 pm

It appears to me to are deleting some items from the train In the third to last photo. Why is that?

Yes, it overproduces items by a small margin, and removes the excess at the unloading station, to keep balancing easier. The removed items are directly voided in this prototype, but conceptually can be transported and consumed by an under-supplied base.

ptx0 wrote: ↑

Sun Mar 21, 2021 10:33 pm

i have trouble buying that this works at all, considering i've tried it with Bob's mods and train-to-train-to-train builds where I use beacons and productivity modules and every time things go out of balance because you can't ensure each assembler has the same level of productivity boost, as different number of overlapping beacon ranges affect each.

can you at least show the train schedule? you mentioned it, but didn't give any hints as to what it looks like.

ah, nevermind. you're using bots. that is an impure setup. it will not go out of balance because the bots don't let it.

train schedule looks like this, nothing fancy. Balancing it is indeed quite tricky. I overproduce a little bit and remove the excess to make it easier.

Do you set item filters in the wagons?
I think that's the only way to make it stable.
I tried this concept with cars on belts. The overproduction kills you ultimately. With dedicated cargo space for each item, you greatly reduce your capacity but couldn't think of another way to get rid of overproduction. (Unloading it into trash/void chests felt like cheeting.)

Bauer wrote: ↑

Mon Mar 22, 2021 7:06 am

Do you set item filters in the wagons?
I think that's the only way to make it stable.
I tried this concept with cars on belts. The overproduction kills you ultimately. With dedicated cargo space for each item, you greatly reduce your capacity but couldn't think of another way to get rid of overproduction. (Unloading it into trash/void chests felt like cheeting.)

exactly. i tried this with a train building circuits, and it's a tightwire balance game to try and ensure one stage doesn't load too much of a single item. it is IMPOSSIBLE to get a full train of goods at the very end. you're always going to have contaminated leftovers.

even when recipes produce 1 item that's consumed at the same ratio at the next stop, if you use productivity modules at all, the ratio is thrown off

Bauer wrote: ↑

Mon Mar 22, 2021 7:06 am

Do you set item filters in the wagons?
I think that's the only way to make it stable.
I tried this concept with cars on belts. The overproduction kills you ultimately. With dedicated cargo space for each item, you greatly reduce your capacity but couldn't think of another way to get rid of overproduction. (Unloading it into trash/void chests felt like cheeting.)

No, to balance it, I overproduce and remove the excess.

siiid wrote: ↑

Mon Mar 22, 2021 4:38 pm

No, to balance it, I overproduce and remove the excess.

This is not possible in a normal vanilla game. Every chest will be full at some point. In vanilla, the only way to balance a production line is backpressure. The only way to make this happen is by setting filters (you have to insert the excess somehow back into the process with priority).

...or we need to talk about the word "balance". I define it as "runs stable without interference for a long time". "stable" = "within a limited range of parameters", e.g. constant output of the desired good, e.g. science packs.

If you have to interfer by eliminating chest content, I don't call it "balanced". We could call it "manually balanced".

Bauer wrote: ↑

Tue Mar 23, 2021 12:40 pm

siiid wrote: ↑

Mon Mar 22, 2021 4:38 pm

No, to balance it, I overproduce and remove the excess.

This is not possible in a normal vanilla game. Every chest will be full at some point. In vanilla, the only way to balance a production line is backpressure. The only way to make this happen is by setting filters (you have to insert the excess somehow back into the process with priority).

The excess items can be transported to another base (that is starving a bit) to consume.

Or they can be just deleted. It is also vanilla feasible:

siiid wrote: ↑

Tue Mar 23, 2021 5:44 pm

The excess items can be transported to another base (that is starving a bit) to consume.

so augument your design to demonstrate that

Or they can be just deleted. It is also vanilla feasible:

Screenshot 2021-03-23 104306.jpg

not really vanilla feasible. expansion, other problems will get in the way. not to mention the nature of having to discover or kill a nest until one worm is left... it's not easy.

demonstrate it outside of sandbox mode

not really vanilla feasible. expansion, other problems will get in the way. not to mention the nature of having to discover or kill a nest until one worm is left... it's not easy.

demonstrate it outside of sandbox mode

That is arguing for the sake of arguing.

siiid wrote: ↑

Tue Mar 23, 2021 9:34 pm

not really vanilla feasible. expansion, other problems will get in the way. not to mention the nature of having to discover or kill a nest until one worm is left... it's not easy.

demonstrate it outside of sandbox mode

That is arguing for the sake of arguing.

if so, then so was your suggestion that a worm can be used to 'reliably' delete materials.

people here are trying to see how useful this concept is, if you didn't want challenging feedback i guess you shouldn't have posted it here

ptx0 wrote: ↑

Tue Mar 23, 2021 10:00 pm

siiid wrote: ↑

Tue Mar 23, 2021 9:34 pm

not really vanilla feasible. expansion, other problems will get in the way. not to mention the nature of having to discover or kill a nest until one worm is left... it's not easy.

demonstrate it outside of sandbox mode

That is arguing for the sake of arguing.

if so, then so was your suggestion that a worm can be used to 'reliably' delete materials.

people here are trying to see how useful this concept is, if you didn't want challenging feedback i guess you shouldn't have posted it here

just create a map with pollution off, biter on, biter expansion off, how difficult is it to build this near the edge of a biter nest? you don't need to "kill until 1 left".

The design has this downside, that it depends on a facility, which reliable consumes the overproduction.
That's not unrealistic. You can create a small science block, which consumes the excess with higher prio and produce the rest from scratch.

Not exactly elegant, but other designs have different downsides.

jodokus31 wrote: ↑

Tue Mar 23, 2021 10:36 pm

The design has this downside, that it depends on a facility, which reliable consumes the overproduction.
That's not unrealistic. You can create a small science block, which consumes the excess with higher prio and produce the rest from scratch.

Not exactly elegant, but other designs have different downsides.

Yes, it is overproducing like 2-3%, therefore a science block of same scale as shown above can handle 20-30 copies of the base.