Should logistic bots be necessary?

[sillyfly]

This thought had recently come to mind, with the addition of the new rocket silo in 0.12.
It is well known that taking items from belts has speed limitation in comparison with taking items from chests/assemblers, because of the stack size bonuses. This is the reason for the ubiquitous green circuit build, with wire assemblers leading directly into circuit assemblers - putting the wires on belts would simply not give enough throughput, because inserters wouldn't be able to pick them up as fast.

With the rocket silo we have similar problems - It takes thousands of ingredients, while taking a very short time (30 seconds) to construct. Combine this with the fact that the largest assembler is still 3x3, you can never get enough items with belts and inserters to make it at full speed.
The same happens for satellite - it needs hundreds of various components, and only takes 3 seconds to craft.
This leaves the player with two options -
1. Use belts and inserters, and face that they will never make full use (or even close to that!) of their assemblers' time.
2. Use logistic robots, taking from various chests filled by many inserters.


For me, I think it is a bit problematic - I have nothing against logistic robots, but I tend to think of them as quality-of-life mechanism, allowing you perhaps to make simpler builds in tight spaces, but should they be necessary? Are we really to be forced to use them to achieve full efficiency?

I would love it if there was a way to make 100% efficient factories without having to resort to logistic robots - sure, if you want to you can, but you should also be able to void them.

This can be done in various ways - the easier would be either re-balancing recipes, so things taking a lot of ingredients would also take longer to construct (3 seconds for a satellite? Why not 30? And a rocket silo in under a minute? Very impressive). A different approach would be to make bigger assembling machines - make a 4x4 or even 5x5 assembling machines, and have the very resource-intensive recipes use only this.

So, what do you guys think? Anyone even agrees with me that this is a problem?

[ssilk]

With introducing a splitting of logistic network like https://forums.factorio.com/forum/vie ... =67&t=8905 it would be possible to nerf the logistic bots a lot. Or have two types of them: One using only the splitted networks, one only the global. The global are much more expensive then.

[ratchetfreak]

you can have multiple insertes taking of the belt and passing the items through chest up to the assembler to let last one take advantage of the stacksize bonus

[sillyfly]

you can have multiple insertes taking of the belt and passing the items through chest up to the assembler to let last one take advantage of the stacksize bonus

This I could, still - it's pretty hard (maybe impossible?) to have enough chests around a 3x3 assembler, with enough inserters leading into each.
But yes, another way to at least partially solve this would be to have larger chests/storage sites - if I could have a 3x3 storage depot, I can have 9 inserters putting things into it from belts on three sides, then 1-3 inserters taking out of it on the other.
Even better would be a 1x5 or 1x6 elongated chest, so I can have belts going on either side of it with 5 or 6 inserters per side, and have a single inserter on the short side feeding into the assembler, taking advantage of the stack size bonuses.

[Adil]

if I could have a 3x3 storage depot

There are such mods:
https://forums.factorio.com/forum/vie ... 94&t=13312
https://forums.factorio.com/forum/vie ... =43&t=7983
Dytech has a nicer graphic (well, zoomed vanilla chest looks better than tiled one as for me) but comes bundled with a bunch of questionable sh*t.
Of the other mod I know only it exists.

1x5 or 1x6 elongated chest

Those are called cargo vagons.

You can also use ches-inserter pairs instead of belts.

[sillyfly]

if I could have a 3x3 storage depot

There are such mods:
https://forums.factorio.com/forum/vie ... 94&t=13312
https://forums.factorio.com/forum/vie ... =43&t=7983
Dytech has a nicer graphic (well, zoomed vanilla chest looks better than tiled one as for me) but comes bundled with a bunch of questionable sh*t.
Of the other mod I know only it exists.

1x5 or 1x6 elongated chest

Those are called cargo vagons.

You can also use ches-inserter pairs instead of belts.

Yes, but that is hardly the point.
I can use mods to change almost anything in the game.
I'm trying to start a discussion about the balance of the base game.

[Adil]

Well, cargo vagons and chest-inserter pairs were part of vanilla game last time I've checked. And while invaluable for novel layouts, multitile chest do feel kind of weird with them effectively moving around large quantities of items in no time and at no cost.

The mods were listed simply in case you were that desperate for functionality you described.

[bobucles]

Each assembler can have up to 8 fast inserters with a full stack bonus going in. You need one point to extract the item. So how do you do it? Can a player take several belt lines going in, and merge them into chests/cargo cars to max out 8 inserters? It sounds like a fun project to try building this.

[sillyfly]

Why 8? You can have 12 if you need no output, or 11 if you want 1 output.

[Tinyboss]

Each assembler can have up to 8 fast inserters with a full stack bonus going in. You need one point to extract the item. So how do you do it? Can a player take several belt lines going in, and merge them into chests/cargo cars to max out 8 inserters? It sounds like a fun project to try building this.

I made a design using only chests and inserters. It should be able to supply 12 fast inserters at full speed (i.e. each inserter feeding the assembler is supplied by five pulling from the belts). It uses 120 fast inserters and 52 chests. I'm sure it could be done a lot easier with cargo cars, and the belt situation could probably be optimized, but here it is:

[MeduSalem]

...

That has got to be the most ridiculous contraption I've ever seen someone build in Factorio.

I like it because it is so completely awkward.

[sillyfly]

...

That has got to be the most ridiculous contraption I've ever seen someone build in Factorio.

I like it because it is so completely awkward.

It's actually good example, because it shows that in fact you can avoid logistic bots, but at the cost of having a very clunky build.

[Zhab]

I think its cool looking.

[bobucles]

Nicely done, but where's the item output?

[Tinyboss]

That has got to be the most ridiculous contraption I've ever seen someone build in Factorio.

Best compliment I've had all day, thanks!

Nicely done, but where's the item output?

It would have to be by...um...logistic bots. Seriously, though, I thought feeding all 12 was an interesting challenge so I solved that. If you actually wanted to use this (without bots), you could easily remove the chest and its infrastructure from one of the cardinal directions, and do your output there.

[Zhab]

Nicely done, but where's the item output?

Well that was a thought experiment for something that would not need an output. Like the rocket silo for example. It will eat a large number of components and self use the result. No need for output.

Of course the rocket silo is much bigger than an assembler. So why Tinyboss build his contraption around an assembler is beyond me. As an assembler would need an output. So the maximum number of chests would be 11 (one for output). But Tinyboss' design can be easily adapted for 11 input and 1 output.

It could theoretically be adapted to have 11.5 inputs with a long handed inserter pair.

It would have to be by...um...logistic bots.

Last I check, robots cannot take stuff directly from an assembler.

[Torquizze]

What about speed inserters Mk.2 with slot for two modules?

[Linosaurus]

This can be done in various ways - the easier would be either re-balancing recipes, so things taking a lot of ingredients would also take longer to construct (3 seconds for a satellite? Why not 30? And a rocket silo in under a minute? Very impressive). A different approach would be to make bigger assembling machines - make a 4x4 or even 5x5 assembling machines, and have the very resource-intensive recipes use only this.

From a game balance point of view there already are a bunch of items that you'll never need to craft at full speed. At least until way later in the game than winning. And it's nice to be able to hand craft them quickly.

For example power poles, locomotives, tanks, radar stations, express belts, construction robots (maybe) - these take 0.5s to craft.

So, what do you guys think? Anyone even agrees with me that this is a problem?

I don't think it is a problem.

It's of course a perfectly valid playstyle to want to design factories so the full production capacity can be used, but I don't think the game should be balanced around it.

--
In the name of awkward solutions... Here's a design for feeding lots of items into one side of a rocket silo (which I think is 5x5? Not using latest version yet). It's awkward and expensive-as-hell, but fairly easy to expand to bigger buildings I think.

Factorio_10LongInserter_20_belts.jpg (211.04 KiB) Viewed 9958 times

[Tinyboss]

On the one hand, each inserter is only getting fed by two belt inserters, so that's not very efficient--there should be five each for maximum exploitation of the stack size bonus.

On the other hand, you've pointed out that 24 inserters can be aimed at a single assembler! Looks like another challenge.

Edit: Challenge accepted.

Here's 24 inserters feeding a single assembler with maximum throughput using only belts. It uses 120 fast inserters, 184 long inserters, and 184 chests.



Here's one side of it, of course they're all identical:


It turns out that using long-handed inserters not only increased the number of inserters that can feed the assembler, but it also made it easier to design the whole thing.

[bobucles]

If the chest has room for two inserters, why not use two inserters? Seems a waste of space to not max out the goods.