Latest science factory (feedback requested)

[Zhab]

Hi everyone. Still new to this game, but I feel like I have learned a lot since my first try. This is my latest science factory design.

(Click image for full size)


My objective was still the same. Make 20+ of any packs in 30 seconds or less. After my first try, I learned that 20 was not such a great number of labs to have and could waste packs depending on how many steps a research have. However, this is no longer a problem in 0.12.

Notable improvement since first try:
1) Much better use of space.
2) Cleaner looking
3) Drastically reduce the size of science pack loop. Only a dozen or so of stockpiled packs per lab.
4) Steel smelted on site.
5) No extra copper cable assembly.

In case of a shortage of iron, the factory will naturally favor red pack first, then green and then blue. So you can research lower tier research until your iron shortage is fixed. This also allow you to build just the parts you need for each colors of packs as you go along and progress in the tech tree and new colors of packs become required.

Good or bad, I want to know what you guys think about it.

[RoddyVR]

Is one side of a belt really enough to feed all those circuit producers? its a blue belt, so i'm no sure, could be (i only ever build sicence factory with yellow belts).

All in all i'd say looks nice, but very hard to judge when its backed up. empty all the belts, and then restart and run the lab for 5 minutes (with research going). Then take a screen shot/ That screen shot will tell a much more detailed story about your factory.

[XyLe]

i see you only have 1 solar panel in there, that's probably not gonna be enough to power all those machines though. but maybe if you take the modules out...

[Talguy]

i see you only have 1 solar panel in there, that's probably not gonna be enough to power all those machines though. but maybe if you take the modules out...

This looks awesome.

[Gandalf]

Reason it's good:
Very compact, looks cool, I assume you made sure all the ratios are (close to) perfect. So that's cool.

Reason it's not amazing:
As with so many of the mini-challenges in Factorio, it doesn't solve a real problem. You made it very compact, which is a real achievement, but there is usually no need to do that and it makes the whole thing very complicated to change/expand.
It's not so much a criticism of your build but rather of the game itself; I think there should be challenge maps with limited space where designs like this are actually needed.

The bit with the shameless self promotion: I just made this which uses the interesting concept of mixed belts. Since you have a hand for compact layouts I'd love to see if you could build a complete science production using mixed belts? That would be even more pointlessly complicated and fun!

[Xterminator]

This really interesting. I have really only ever built the same type of science setup and I find it really interesting seeing designs like this.
Here are my thoughts on it.
I like it because it is compact, which can be quite hard to do well, and it looks pretty damn cool. It also looks like it works quite well for the current demand.

Here is what I don't really like. Pretty much the same thing Gandalf said. The one downside to these compact designs like this is that they are super hard to expand any further without tearing down a bunch of things. Granted, you can upgrade the assemblers to make them faster, but past that it becomes quite hard to expand.

All in all I would say good design! It does what it is supposed to and looks cool. Only major downside is lack of expandability.

[vampiricdust]

Reason it's good:
Very compact, looks cool, I assume you made sure all the ratios are (close to) perfect. So that's cool.

Reason it's not amazing:
As with so many of the mini-challenges in Factorio, it doesn't solve a real problem. You made it very compact, which is a real achievement, but there is usually no need to do that and it makes the whole thing very complicated to change/expand.
It's not so much a criticism of your build but rather of the game itself; I think there should be challenge maps with limited space where designs like this are actually needed.

The bit with the shameless self promotion: I just made this which uses the interesting concept of mixed belts. Since you have a hand for compact layouts I'd love to see if you could build a complete science production using mixed belts? That would be even more pointlessly complicated and fun!

There's a challenge scenario called Tight Spot where you have to make really compact & efficient builds to earn money/points.

[Zhab]

Is one side of a belt really enough to feed all those circuit producers? its a blue belt, so i'm no sure, could be (i only ever build sicence factory with yellow belts).

All in all i'd say looks nice, but very hard to judge when its backed up. empty all the belts, and then restart and run the lab for 5 minutes (with research going). Then take a screen shot/ That screen shot will tell a much more detailed story about your factory.

You have a good eye sir. Indeed a blue belt is not enough to handle the production of 6 copper cable assemblers. However that is the wrong screen shot (wasn't the final version). I've edited my post to show the proper screenshot. A double red belt isn't quite enough to handle 5 assembly running at 1.25 crafting speed. But the truth is that I don't need maximal speed. You can run this with blue assemblers with 2 speed modules in it for a crafting speed of 0.95. But a yellow assembler use less power. I could use only 5 assemblers once I get yellows. But you don't start with yellows and I wanted to show that there is meant to be 6 assemblers for a long while in the screenshot.

Also, the first gear assembler was interfering with the output of the second one. Occasionally delaying output. Also, the fast inserter assembler was not getting quite enough iron to match the speed of the yellow inserter inserting inserters. All is fix in this version.

i see you only have 1 solar panel in there, that's probably not gonna be enough to power all those machines though. but maybe if you take the modules out...

Actually, I have this mod that makes solar panels 1000 times stronger. No I don't. But seriously, the panel is just my way of transforming useless space into useful space. It is not like panels need to be in specific areas to be useful.

Reason it's good:
Very compact, looks cool, I assume you made sure all the ratios are (close to) perfect. So that's cool.

Reason it's not amazing:
As with so many of the mini-challenges in Factorio, it doesn't solve a real problem. You made it very compact, which is a real achievement, but there is usually no need to do that and it makes the whole thing very complicated to change/expand.
It's not so much a criticism of your build but rather of the game itself; I think there should be challenge maps with limited space where designs like this are actually needed.

The bit with the shameless self promotion: I just made this which uses the interesting concept of mixed belts. Since you have a hand for compact layouts I'd love to see if you could build a complete science production using mixed belts? That would be even more pointlessly complicated and fun!

Well this is meant to be end game. Do you really NEED more than 20 research labs ? I mean this need a full red belt of iron to keep it running smoothly. That is a lot of iron to dedicate to just research. I was under the impression that most people went with 10 labs. But maybe I am mistaken. Of course I have heard about that 60 seconds missile defense (*cough* rocket silo) type of play through but... that is hardly standard now is it ?

But the point is, this was built with the idea that it would never ever need to be expanded. I personally feel like 20 research labs are more than enough. But hey ! If you need 40, just build a second one of these and you are all set. So I guess this can be expanded... in chunks of 20 labs.

But you are not meant to start the game with 20 labs. You don't have the resources for that anyways. You just build the red science part and 5 labs or so. Then you build the green part and upgrade to 10 labs or so. Then you construct the huge blue science area and upgrade to the full 20 labs. It is a design that is meant to be built and most importantly used as you play along. It is fully usable while incomplete. Most of what you see is there just for blue science.

As for your design, it is an interesting proof of concept. I had something similar in mind for my factory. I'm trying real hard to avoid the "main bus" or "robots do everything" strategies that all the cool kids are doing.

This really interesting. I have really only ever built the same type of science setup and I find it really interesting seeing designs like this.
Here are my thoughts on it.
I like it because it is compact, which can be quite hard to do well, and it looks pretty damn cool. It also looks like it works quite well for the current demand.

Here is what I don't really like. Pretty much the same thing Gandalf said. The one downside to these compact designs like this is that they are super hard to expand any further without tearing down a bunch of things. Granted, you can upgrade the assemblers to make them faster, but past that it becomes quite hard to expand.

All in all I would say good design! It does what it is supposed to and looks cool. Only major downside is lack of expandability.

Woot, a youtuber commenting on my design. I've watched a lot of your videos good sir.

Again... does 20 research labs really need to be expanded ? How many labs to you usually end up using ? Do people usually dedicate several full red belts of resources to research alone and to hell with everything else ? Seems like madness to me. I mean... there isn't that much stuff to research for extreme amount of labs to be worthwhile is there ?

But because this makes very good use of space and is very close to ratios as possible, if you need more labs you can just "expand" by making a blueprint of the layout and pasting it over. 2 = 40 labs. 3 = 60 labs. Etc. In my opinion, you'll end up with something more compact and neater looking than an irregular shape expanding up and down from the main bus.

I guess it is comparable to the famous Arumba furnaces layout. A some point it is better to build an other one on the side than continue stretching the current one.

It is also comparable to solar panels modules. You expand by copy pasting the whole thing.

I guess this is a clash between the modular factory strategy and the main bus strategy.

Thank you all for your feedback.

[MalContentFL]

I really like the setup. My blue science pack production always turn into a mess. I've seen people give a blueprint string for their designs. 1. Can we have one for your design? 2. How do we use them?

[Zhab]

I really like the setup. My blue science pack production always turn into a mess. I've seen people give a blueprint string for their designs. 1. Can we have one for your design? 2. How do we use them?

I.... hmmm... I don't know either lol. My factorio experience have been very solitary so far. I never needed to share blueprints before or even considered doing so. I will have to look into that.

[DerivePi]

Checked the numbers using the 8 assembler 3 blue science as the desired rate (1.2 red, green and blue science produced per second). Here are the bottlenecks:
- need over 23 iron plates per second - express belt
- need more than 7 steel furnaces
- need 4 assembler 3's for circuit production - need more than 4 assembler 2s with speed mod 1s.
- need almost a full express belt of copper cables - very difficult to accomplish with just inserters
- petro production will need 5 advanced refineries and a full complement of chemical crackers

At 1 set of science per second, you almost have the right balance.

Thank you for sharing.

[Zhab]

Checked the numbers using the 8 assembler 3 blue science as the desired rate (1.2 red, green and blue science produced per second). Here are the bottlenecks:
- need over 23 iron plates per second - express belt
- need more than 7 steel furnaces
- need 4 assembler 3's for circuit production - need more than 4 assembler 2s with speed mod 1s.
- need almost a full express belt of copper cables - very difficult to accomplish with just inserters
- petro production will need 5 advanced refineries and a full complement of chemical crackers

At 1 set of science per second, you almost have the right balance.

Thank you for sharing.

1 science per second. Well that is not what I was gunning for I have 20+ science pack in 30 seconds or less. In other words, at lest 2/3 blue science per seconds. Which is significantly less.

I did my initial calculation with a crafting speed of 1. However a crafting speed of one is actually actually hard to get near to in the game. I can manage with blue assemblers with 2 speed modules in it. However when it comes to power usage and pollution it is actually better to use a yellow assembler at that point. Yellow assemblers are faster than I need, but I don't have to run them at full speed. as long as loop is stuffed with science packs and science gets done by 20 science labs running 24/7 thanks to a compact layout, I'm happy.

PS: You sound way more experienced with the finer aspects of the game then me.