Factorio Forums

www.factorio.com
https://forums.factorio.com/

Add "quality priority" to inserters

https://forums.factorio.com/viewtopic.php?t=117635

Page 2 of 2

Re: Add "quality priority" to inserters

Posted: Sun Apr 19, 2026 8:19 am
by terradus
just Open a calculator, pick a finished product, and then imagine switching that end product mid-process. If you build a static setup without the ability to switch, your factory ends up oversized and idling. Meanwhile, you know there are other players who might need that space, and rebuilding would take precious time you don't have.
Now, let's sum up all the belts needed for every intermediate production step. For example: to produce 55.379 modules per minute, you need raw materials. One red belt of copper ore will be 99.99% full, another red belt of iron ore at 46.15%, and coal at 20.65% (excluding oil and water pipes). That totals 166.79% in 'belt capacity' at the input.
However, if we sum all the belts used throughout the entire production flow down to the final product (excluding the raw ore inputs), it reaches 438.43%! And remember, the goal is still to switch to a different end product without rebuilding.
Since you need to separate resources by quality onto different belts, multiply that 438.43% by 3 (for the early-game quality tiers). You get 1315.29%. Compare that to the 166.79% at the input. Think about how much space you’d save if these items weren't on belts at all, but stayed in a chest! From there, through automated switching, they could be processed into the final product.
Consider the challenges when your goal is to squeeze every bit of value out of limited space and resources. You can't just spam productivity modules—you might lack the machines or the raw materials. Your only chance is to maximize output through Quality Modules and balance the entire factory based on mathematical probability. All of this is to get the maximum result from minimum space.
In this context, 'priority' becomes a vital concept. You can't automatically switch all filter inserters to 'any quality module' because the developers didn't include an 'any quality' signal or a 'quality priority' parameter. This omission breaks the standard in-game logic you’re used to.

I'm not asking for a 'win button.' I’m simply suggesting the completion of a logical feature that you’d expect to already be there—only to find out, once you start building your dream factory, that the functionality you counted on is actually missing.

I’m not trying to reshape the game to my liking or begging the devs for favors. It’s just that some things are obvious: when everything operates within one paradigm, but one specific element falls out of line, you immediately realize it’s either an error or an oversight.
My main point is this: you can manually set a filter inserter to 'Any Quality Module,' but you can't send that same thing as a signal. That is a flaw; it shouldn't be this way. I probably should have just gone straight to the bug reports.

P.S. I’m using a translator, so I might miss some nuances in tone. I’ve tried my best to be polite and respectful, so please excuse me if anything came across the wrong way. I have huge respect for all of you, and the game is absolutely amazing! But there is always room for improvement. Thanks in advance, everyone. Best regards.

Re: Add "quality priority" to inserters

Posted: Sun Apr 19, 2026 8:21 pm
by mmmPI
terradus wrote: Sun Apr 19, 2026 8:19 am just Open a calculator, pick a finished product, and then imagine switching that end product mid-process.
No calculator , but challenge accepted ! x)

Here is my version :
module factory.png
module factory.png (3.39 MiB) Viewed 574 times


I tried to make it similar as you for module tier 1, and about the same size and inputs, one can change which module type is produced from the constant combinator. For now it recycle everything that is not at least "rare", but it's easy to adapt to keep all the modules, i was not sure how your machine function, i think yours is using quality ores as input, not my version because it's easier without, and there are also other differences probably. I used a rare heat tower, and half of the build is static, because it was easier for me to make it small this way. I used direct insertion, belt, chest and even train wagon to deal with the volume of intermediate material and made some loops to avoid problems when receipe changes. I also used speed and productivity modules in about half the machines, the static ones, because i like them and i didn't understand why you only used quality.
terradus wrote: Sun Apr 19, 2026 8:19 am I'm not asking for a 'win button.' I’m simply suggesting the completion of a logical feature that you’d expect to already be there—only to find out, once you start building your dream factory, that the functionality you counted on is actually missing.
When i tried to build previous blueprint, looking for hidden rocks ^^ , i felt this way for the other suggestion : viewtopic.php?t=126836, not so much for "quality priority", but that is just my opinion. If there are plenty use-case people post, it makes more arguments about gameplay than just asking for a feature. Sometimes some use-cases are because player are missing something, sometimes there are other solutions, sometimes not, i don't know, it's the devs who decide. I think it helps them understand what players are attempting to do when they ask for a feature if they also post their setup, maybe it is a beginner and more tooltips for existing features are the solutions sometimes, but other times it's a player who know very well the game and don't need more tooltip and there are many times it's not so clear, i feel sharing blueprints is better than trying to put label on people :)

This is different than the current suggestion, it doesn't work in game, but i missed it :
quality signal as filter for inserter.jpg
quality signal as filter for inserter.jpg (65.63 KiB) Viewed 574 times
terradus wrote: Sun Apr 19, 2026 8:19 am P.S. I’m using a translator, so I might miss some nuances in tone. I’ve tried my best to be polite and respectful, so please excuse me if anything came across the wrong way. I have huge respect for all of you, and the game is absolutely amazing! But there is always room for improvement. Thanks in advance, everyone. Best regards.
I have been impatient when asking if you used a robot to answer , maybe disrespectful, sorry if that is the case, the situation felt bizarre and it made me laugh, i've tried to make more effort to write a proper english for translation this time, hopefully :)

Re: Add "quality priority" to inserters

Posted: Sun Apr 19, 2026 10:11 pm
by terradus
mmmPI wrote: Sun Apr 19, 2026 8:21 pm
No calculator , but challenge accepted ! x)
A challenge should have rules, and the rule here is to gain quality at every single stage. I’ve built setups like yours before, but this is something fundamentally different. My goal is to extract quality at every production step and feed it into the next. Try to rework everything: every single assembler must be filled with Quality Modules and nothing else. And the whole system must be controlled by a constant combinator that can switch the entire production line. On top of that, the goal is to minimize both downtime and belt clutter (clogging). Also, keep in mind that you are destroying the production process at the end (wasting system uptime, energy, and 75% of all resources) instead of enriching the ore, which is much more efficient because it saves the entire system's workload.

Unfortunately, we’re unlikely to understand each other because you’re focusing on just one piece of the puzzle, whereas to be on the same page, you need to look at the whole picture. Here are the constraints:
1) All modules must be Quality only.
2) Minimum footprint (space efficiency).
3) No recycling of finished products (only ore recycling is allowed as a baseline). This means if you have ore recycling working for balance, you are not permitted to recycle finished production. In other words, ore recycling alone must provide the balance for quality resources. you can recycle finished products if ore recycle is available and works for balance.
4) Zero downtime.
5) The system must be reconfigurable to any other end product at any moment via a combinator — without manually touching the inserters (or without rebuild).
6) No failures and zero maintenance required.
Until you see how these points work together, we're talking about different things.

Re: Add "quality priority" to inserters

Posted: Mon Apr 20, 2026 12:18 am
by mmmPI
terradus wrote: Sun Apr 19, 2026 10:11 pm A challenge should have rules
I am not very good at respecting the rules, for me it was :
terradus wrote: Sun Apr 19, 2026 8:19 am just Open a calculator, pick a finished product, and then imagine switching that end product mid-process.

For example: to produce 55.379 modules per minute,
But i didn't open a calculator, so it didn't start very well, i added the rule : not burn coal and for the production i just tried to use about the same number of assembly machines as you used. But i am unable to use your blueprint so i didn't know how much is recycled and why. It is possible to recycle only the excess above 55 per minute instead of almost everything with a few extra combinators blocking a splitter in the top right corner, i was too lazy to add.
2) Minimum footprint (space efficiency).
5) The system must be reconfigurable to any other end product at any moment via a combinator — without manually touching the inserters (or without rebuild).
6) No failures and zero maintenance required.
I also did those but for rule 5, i noticed your blueprint is only for tier 1 module, if you want tier 2, you need to reconstruct, so i thought : it will be easy to limit the downtime and logic of the machines that produce intermediate, because it is always the same. So i made a part static, since it doesn't use quality module, the ratio are not expected to change and only minimal intermediate product variety.

Even with the other rules it doesn't specific that ALL machines must be switching recipe.
1) All modules must be Quality only.
3) No recycling of finished products (only ore recycling is allowed as a baseline). This means if you have ore recycling working for balance, you are not permitted to recycle finished production. In other words, ore recycling alone must provide the balance for quality resources. you can recycle finished products if ore recycle is available and works for balance.
4) Zero downtime.
The rule 3 is still confusing for me. I feel this is your challenge for the mountain fortress scenario, your decision to use that strategy. I understand the use-case i feel, you want to use quality module everywhere, and no beacons, no robots and only tier 1 module, no blue belts, many things, i understand why or i imagine it is because of the available tech and ressources when you play. I feel it would be easier if the other suggestion was implemented, to be able to set a quality level as filter for inserter with signal. I don't understand still how the "quality priority" is supposed to help you. I don't see how it would make things easier to adapt my build to match all the rules, i would instead prefer to have the possiblity to set a spiltter to filter this :
any quality item in filter splitter.jpg
any quality item in filter splitter.jpg (33.77 KiB) Viewed 510 times
terradus wrote: Sun Apr 19, 2026 10:11 pm Unfortunately, we’re unlikely to understand each other because you’re focusing on just one piece of the puzzle, whereas to be on the same page, you need to look at the whole picture. Here are the constraints:

Until you see how these points work together, we're talking about different things.
It is correct that i try to focus on just one piece now, because i feel part of the rules you explained are caused by the scenario or that different player would choose different methods to solve the challenge like many train wagons or cars instead of chest. And then it could be off topic with the suggestion. It is just my opinion. I want to understand more why "quality priority" is important in your build. Or maybe in other build, maybe a simpler version where it is easier to see it in action. You said in your post earlier that you wanted it to work on belt like "freshness priority", but freshness priority doesn't work on belts, i also want to make sure this is not something overlooked.

Re: Add "quality priority" to inserters

Posted: Mon Apr 20, 2026 11:26 am
by terradus
mmmPI wrote: Mon Apr 20, 2026 12:18 am
terradus wrote: Sun Apr 19, 2026 10:11 pm A challenge should have rules
I am not very good at respecting the rules, for me it was :
terradus wrote: Sun Apr 19, 2026 8:19 am just Open a calculator, pick a finished product, and then imagine switching that end product mid-process.

For example: to produce 55.379 modules per minute,
But i didn't open a calculator, so it didn't start very well, i added the rule : not burn coal and for the production i just tried to use about the same number of assembly machines as you used. But i am unable to use your blueprint so i didn't know how much is recycled and why. It is possible to recycle only the excess above 55 per minute instead of almost everything with a few extra combinators blocking a splitter in the top right corner, i was too lazy to add.
2) Minimum footprint (space efficiency).
5) The system must be reconfigurable to any other end product at any moment via a combinator — without manually touching the inserters (or without rebuild).
6) No failures and zero maintenance required.
I also did those but for rule 5, i noticed your blueprint is only for tier 1 module, if you want tier 2, you need to reconstruct, so i thought : it will be easy to limit the downtime and logic of the machines that produce intermediate, because it is always the same. So i made a part static, since it doesn't use quality module, the ratio are not expected to change and only minimal intermediate product variety.

Even with the other rules it doesn't specific that ALL machines must be switching recipe.
1) All modules must be Quality only.
3) No recycling of finished products (only ore recycling is allowed as a baseline). This means if you have ore recycling working for balance, you are not permitted to recycle finished production. In other words, ore recycling alone must provide the balance for quality resources. you can recycle finished products if ore recycle is available and works for balance.
4) Zero downtime.
The rule 3 is still confusing for me. I feel this is your challenge for the mountain fortress scenario, your decision to use that strategy. I understand the use-case i feel, you want to use quality module everywhere, and no beacons, no robots and only tier 1 module, no blue belts, many things, i understand why or i imagine it is because of the available tech and ressources when you play. I feel it would be easier if the other suggestion was implemented, to be able to set a quality level as filter for inserter with signal. I don't understand still how the "quality priority" is supposed to help you. I don't see how it would make things easier to adapt my build to match all the rules, i would instead prefer to have the possiblity to set a spiltter to filter this :

any quality item in filter splitter.jpg

terradus wrote: Sun Apr 19, 2026 10:11 pm Unfortunately, we’re unlikely to understand each other because you’re focusing on just one piece of the puzzle, whereas to be on the same page, you need to look at the whole picture. Here are the constraints:

Until you see how these points work together, we're talking about different things.
It is correct that i try to focus on just one piece now, because i feel part of the rules you explained are caused by the scenario or that different player would choose different methods to solve the challenge like many train wagons or cars instead of chest. And then it could be off topic with the suggestion. It is just my opinion. I want to understand more why "quality priority" is important in your build. Or maybe in other build, maybe a simpler version where it is easier to see it in action. You said in your post earlier that you wanted it to work on belt like "freshness priority", but freshness priority doesn't work on belts, i also want to make sure this is not something overlooked.
To improve clarity, I’m currently refining my factory and planning to include a detailed operating manual. This will visually demonstrate its mechanics and allow for comparison with similar builds. It will also help me finalize the design and calculate the exact footprint of the logic circuitry, which was necessitated by the lack of a 'quality priority' parameter in inserters.
But even setting 'Mount Fortress' aside, I plan to use this blueprint in other playthroughs as well (including vanilla). The concept remains virtually the same — essentially, it’s an Omni-factory. This term was coined by someone in the community, either on Reddit or the forums, I don't recall exactly. The idea of an Omni-factory is to produce any item list configured in a constant combinator. I’ve tried many different 'Omni' designs under various names, but none satisfied me because they all had significant drawbacks.
Notably, I haven't seen a single Omni-factory that successfully integrates production lists of varying qualities while using quality modules. This is a major pain point that I’ve experienced firsthand. To understand why this is such an issue, consider this: you can manually set 5 'Any Quality' filters on an inserter (see
04-20-2026, 14-30-09.png
04-20-2026, 14-30-09.png (97.46 KiB) Viewed 432 times
), but achieving the same functionality via circuit logic is impossible. This is because the signal analogs take up way more space (25 slots instead of 5). Sure, you could design logic to pick the right signal out of those 25, but it consumes physical space and 'developer brainpower' for very little return. This inefficiency kills the profit of developing Omni-factories in the first place.
Think about it: is it right that you can manually set 5 'Any Quality' filters, but such signals don't exist in the circuit logic? I believe this is a flaw. Soon, I’ll share my blueprint with a detailed description to make everything clear.

Re: Add "quality priority" to inserters

Posted: Mon Apr 20, 2026 1:04 pm
by Tertius
terradus wrote: Mon Apr 20, 2026 11:26 am Notably, I haven't seen a single Omni-factory that successfully integrates production lists of varying qualities while using quality modules.
This is a problem for you? All of my "omni" upcycling factories work this way. Upcycling factories, not production factories. I developed several of these, because the one and only omni factory doesn't exist. You need at least one for every manufacturing machine: assembling machine, foundry, electromagnetic plant, biochamber, cryogenic plant. And for optimization purposes, you might need specific designs for specific items.

And it doesn't work for a factory that doesn't do upcycling the final product but directly craft one specific quality, usually legendary. You create legendary ingredients wherever you want, then use a classic autofactory that's doing the same as it does for normal components: load ingredients and craft. No recycling.

How I build these omni upcycling factories in general:
  • dedicated machines for every quality, I have usually 4 for normal, 1-2 for uncommon, 1 for rare, 1 for epic, 1 for legendary. Will also work for just 1 machine that switches not only between items but also between every quality, but I didn't do this, because it's too slow.
  • a list of desired items defined in a constant combinator
  • a list of ingredients is created for every quality, i. e. 5 recipes. Created by setting the recipe of the item to craft, reading ingredients and create 5 ingredient sets with 5 selector combinators with quality transfer
  • a list of to-keep output items is created, contains up to 5 items, one for each quality
  • a list of to-recycle output items is created, contains up to 5 items, one for each quality
  • one long sushi belt visiting every machine that carries a collection of all required ingredients
  • one long sushi belt visiting every machine that carries the output items away
  • a bunch of recyclers being fed from the output item sushi belt and outputting items back to the ingredient sushi belt
Now to the filters:
  • Since we have the list of to-keep items we set these in an inserter whitelist filter placed at the output sushi belt. These items are being extracted and stored as the result we want to have. Since it's an upcycling factory, items of all quality will eventually get here, so we're able to extract every possible quality end product.
  • The same list is set as blacklist filter in the inserters moving items from the output sushi belt into the recyclers.
There are more filters, because the ingredient sushi belt needs active balance.
To keep it supplied, we create a set of wanted items on the sushi belt. It's something like 5 times the normal ingredients + 3 times the uncommon ingredients + 2 times the rare ingredients + epic ingredients + legendary ingredients, computed from the 5 ingredient sets. We read and subtract the current sushi belt content from this list of wanted items and request the missing items in a requester chest and move all items from it on the sushi belt.
And we have to remove surplus items to keep the sushi belt from overfilling and stalling. We compute the inverse of the above list, subtract some constant value for some leeway and set this as filter in an inserter that will remove surplus items and put them into an active provider chest.

In general, that's it.

I also have a completely bot based omni upcycling factory variant. No sushi belts, instead requester chests and active provider chests. Instead of filling the sushi belt, ingredients can directly be requested for the corresponding machine. Simpler layout but all the ingredients of all quality are distributed over the whole logistic network. To avoid, needs separate logistic networks.

Example Aquilo specific items, is able to upcycle everything that can be crafted by a cryogenic plant: (bot based)
04-20-2026, 14-45-31.png
04-20-2026, 14-45-31.png (3.09 MiB) Viewed 414 times
Example Vulcanus specific items, is able to upcycle everything that can be crafted by a foundry: (bot based)
[the mining drill ingredient factory to the right is strictly normal quality, it's supposed any base ingredients are normal and being supplied from outside the upcycling factory]
04-20-2026, 14-46-42.png
04-20-2026, 14-46-42.png (2.7 MiB) Viewed 414 times
Example module factory (exclusively for modules, it's so big because it's intended for starting with quality so every machine is just normal quality, because that's all we have at the start. Able to craft all 4 modules eff, prod, speed, qual) (sushi belt based):
04-20-2026, 14-49-35.png
04-20-2026, 14-49-35.png (8.56 MiB) Viewed 414 times

Re: Add "quality priority" to inserters

Posted: Mon Apr 20, 2026 1:13 pm
by terradus
Tertius wrote: Mon Apr 20, 2026 1:04 pm
This is a problem for you?
This is another variation entirely, not the one I’m talking about. That’s likely why there’s a misunderstanding, though I understand your point perfectly. I realize I’ve written a lot here and it can be a lot to go through, so no worries. Best regards.

Re: Add "quality priority" to inserters

Posted: Mon Apr 20, 2026 1:30 pm
by Tertius
terradus wrote: Mon Apr 20, 2026 1:13 pm
Tertius wrote: Mon Apr 20, 2026 1:04 pm
This is a problem for you?
This is another variation entirely, not the one I’m talking about. That’s likely why there’s a misunderstanding, though I understand your point perfectly. I realize I’ve written a lot here and it can be a lot to go through, so no worries. Best regards.
I tried to find a reason why the "quality priority" from the OP is useful, and I found none. The best explanation or explained motivation was your post, so I answered it. It's actually an answer why I don't see a reason for the OP request.

If there was a "quality priority" option similar to the "spoilage priority" - what would I be able to do with it? I would be able to pull the item with the highest or lowest quality from a chest or a machine (an inventory is required). But I cannot influence which quality this would be. It's a random quality, because any quality can be in the chest. In the end, I'm pulling some random item, and I don't see how this can be useful.

I need a specific quality, or I need to identify the quality I pulled to make it useful. So I can just read the chest, filter by quality, and pick the desired item. In this case I'm pulling an identified item and I know exactly what I can do with it.

It's also already possible to kind of prioritize quality, because in item order the higher quality has the higher priority. So I can read a chest, filter eligible items with EACH and output ANY (the green wildcard). The ANY will pick the item with the highest priority, and if you filtered the same item but all its different qualities, this will pick the item with the highest quality.

Re: Add "quality priority" to inserters

Posted: Tue Apr 21, 2026 11:06 am
by mmmPI
terradus wrote: Mon Apr 20, 2026 11:26 am But even setting 'Mount Fortress' aside, I plan to use this blueprint in other playthroughs as well (including vanilla). The concept remains virtually the same — essentially, it’s an Omni-factory. This term was coined by someone in the community, either on Reddit or the forums, I don't recall exactly. The idea of an Omni-factory is to produce any item list configured in a constant combinator. I’ve tried many different 'Omni' designs under various names, but none satisfied me because they all had significant drawbacks.

Notably, I haven't seen a single Omni-factory that successfully integrates production lists of varying qualities while using quality modules. This is a major pain point that I’ve experienced firsthand.
hehe i built a few of those and here it's a bit outside of the scope of the suggestion i think but my concern would be then that in a normal game if you deal with "rare" ore , you may want to use productvity module to turn them into circuit, rather than quality module, because they cannot receive anymore quality upgrade but in your omni factory you have used quality module everywhere. Would this qualify as a significant drawback ? a tiny one ? it's not important :lol: But i understand the reasons to make your own when there is someting you would prefer different than what exist.

Also it is much easier to integrate quality at every step if you use robots and also maybe people would use electromagnetic plant for their module omni factory.
terradus wrote: Mon Apr 20, 2026 11:26 am To understand why this is such an issue, consider this: you can manually set 5 'Any Quality' filters on an inserter but achieving the same functionality via circuit logic is impossible.
This i agree it is my conclusion for now and i was thinking about that, i feel in your use case maybe it is possible to read the machine's ingredient and do some quality transfer. This is not possible when the use case is a train station. It is just a thought.
Tertius wrote: Mon Apr 20, 2026 1:30 pm If there was a "quality priority" option similar to the "spoilage priority" - what would I be able to do with it? I would be able to pull the item with the highest or lowest quality from a chest or a machine (an inventory is required). But I cannot influence which quality this would be. It's a random quality, because any quality can be in the chest. In the end, I'm pulling some random item, and I don't see how this can be useful.
This is also my concern, i feel "quality priority" isn't like "freshness priority", it's too different to use the consistency argument for adding the settings. For the use cases mentionned, the ability to set generic or item related quality filter from circuit would be more suited.
Tertius wrote: Mon Apr 20, 2026 1:30 pm It's also already possible to kind of prioritize quality, because in item order the higher quality has the higher priority. So I can read a chest, filter eligible items with EACH and output ANY (the green wildcard). The ANY will pick the item with the highest priority, and if you filtered the same item but all its different qualities, this will pick the item with the highest quality.
That's a nice trick ! Unfortunatly it doesn't cover the cases i found the most frustrating : splitters that have only 1 filter slot, where you can't read the inventory to pick the selected thing that will pass thru, if you want to change the item, and you are doing something quality related, it quickly becomes annoying because any filter set by circuit will be of NORMAL quality.

The use case would be to try to do "quality at every step", unlike the setup we posted which are recycling loops of end products, i usually use robots and "trash unrequested" or active prodvider to deal with the intermediates and different quality level.
quality at every step.jpg
quality at every step.jpg (374.42 KiB) Viewed 277 times
If you try to do something like this because you have iron ore or plate with different quality level as input, but with belts instead of robots, and dynamic recipe for every module and to avoid machine idling as much as possible.
Then, to me, the ability to set quality filter by circuits feels missing. Maybe i'm not seeing how "quality priority" could be used, so don't "miss" it.

Re: Add "quality priority" to inserters

Posted: Wed Apr 22, 2026 5:01 am
by terradus
one more reason
04-22-2026, 08-01-13.png
04-22-2026, 08-01-13.png (748.75 KiB) Viewed 154 times

Re: Add "quality priority" to inserters

Posted: Wed Apr 22, 2026 5:09 am
by mmmPI
terradus wrote: Wed Apr 22, 2026 5:01 am one more reason 04-22-2026, 08-01-13.png
this setup has no room to remove spoilage, it will break down anyway :( But yeah for science packs "quality priority" would make sense.
All times are UTC
Page 2 of 2

Powered by phpBB® Forum Software © phpBB Limited