Can Logistics robots replace conveyor belts and inserters for small areas?

[InfestedAdam]

I think I finally understand how to use these robots. I'd imagine for long distances conveyor belts or trains are the go-to items but for short distances or small areas, especially if I can't get the "perfect" layout between convenyor belts, inserters, etc are logistic robots the best approach? If properly setup are they fast, as fast, or slower compared to convenyor belts and inserters?

Comments are appreciated,

Thank you

[The Chaotic Coder]

Logistic networks are much simpler to use and more flexible than trying to route belts with 3 or 4 components to a row of assemblers. Each machine just needs 2 inserters, one grabbing components from a requester chest and the other putting the product into a provider chest.

You can even share chests between pairs of machines to lower cost:

http://steamcommunity.com/sharedfiles/filedetails/?id=880378151

Over long distances, trains are much faster, cheaper, better in almost every way than bots or belts (plus I personally thing they're awesome!) but moving stuff around within a factory, bots are definitely the way to go (if you can spare the power for them, of course)

Currently, robot networks have a much lower CPU load than belts, so if you're planning on making a really huge factory they can help with game performance (although I've heard that the 0.15 update will have optimizations for belts, so it remains to be seen which one will be better for performance after that)

[InfestedAdam]

Thank you for the response. I got my system setup with bots making bots and already notice the difference in not having to have the perfect layout of belts, assemblers, and inserter. I think I'll have the construction bots pick up everything and redesign the place.

[KatherineOfSky]

Originally posted by jchardin64:

Logistic networks are much simpler to use and more flexible than trying to route belts with 3 or 4 components to a row of assemblers.

I don't agree that they are simpler to use. If you are not careful on adjusting your requests, you can easily have some machines starved because of prioritization of higher request amounts.

Bots can be useful, especially if you want to transfer items from annoying locations, or use stuff you wouldn't otherwise put on the bus. However, I definitely prefer a bus setup because I can much more easily control the prioritization and throughput of intermediate products.

If you use bots, you must make sure you have sufficient production to feed everything.

[!?!]

If properly setup, you have infinite, self-sorting throughput.
Of course belts can't compete.
If improperly setup, nothing gets done.

[haessliches Frettchen]

for the main items that often needed i prefer belts, beacuse i will not see 3000 flying bots, they are slow and for me it looks better with belts...
for some products that not often needed and which needs mats that just some machines needed i take bots, like solid fuel for trains, flamethrower ammo, alien artefacts for modules...

[The Chaotic Coder]

Originally posted by KatherineOfSky:

Originally posted by jchardin64:

Logistic networks are much simpler to use and more flexible than trying to route belts with 3 or 4 components to a row of assemblers.

I don't agree that they are simpler to use. If you are not careful on adjusting your requests, you can easily have some machines starved because of prioritization of higher request amounts.

What I mean by "simple" is they completely avoid "OK now I have to split off several belts of components from the bus, and figure out how to get all of them through here in a limited space".

Even with a "perfect" bus setup, you are limited in throughput by the amount of resources that blue belts can carry. A furnace collumn can't be more than 72 furnaces long because that's the maximum a blue belt can carry. Just adding more and more belts, you'll always eventually run out of room somewhere.

Originally posted by KatherineOfSky:

Bots can be useful, especially if you want to transfer items from annoying locations, or use stuff you wouldn't otherwise put on the bus. However, I definitely prefer a bus setup because I can much more easily control the prioritization and throughput of intermediate products.

If you use bots, you must make sure you have sufficient production to feed everything.

Not necessarily. You can always put conditions on the inserters to only operate if the amount of material in the network is >X. Use different X for different things to control priority. For example:

--Inserter feeding iron into gear machines only operates if the amount of iron in the system is > 100. The output inserter operates if the amount of gears < 200.
--Inserter feeding iron into pipe machines operates if the amount of iron in the system is > 250. The output inserter operates if the amount of pipes < 400.
--Inserter feeding iron into steel smelters operates if the amount of steel in the system is < 1000 (and it doesn't care about iron supply).

So steel gets first priority until its quota is met. Gears have a higher priority than pipes do, so if there isn't enough iron coming in for everything then pipes will shut down first, followed by gears.

[Kintharo]

I used to make a huge factory with a rotating main bus. The game slows down...

Edit:

I made some screenshots of that. But i am on my android right now, makes to show a link difficult

[Jan]

Every lategame factory should be a robot hive. No belts allowed.

[Kintharo]

Dear jan,

This is your energy bill:
354.605 teraW.
Total cost: 996.658 US dollar.
Pay the full amount to our bank bic/swift account 5368.
We will send the next bill within a hour.

Sincere, your energy company

[!?!]

Originally posted by Jan:

Every lategame factory should be a robot hive. No belts allowed.

I've found that mixing bots with belts to minimize bot travel distance helps a lot with maximizing bot efficiency.

[KatherineOfSky]

Originally posted by jchardin64:

Originally posted by KatherineOfSky:

I don't agree that they are simpler to use. If you are not careful on adjusting your requests, you can easily have some machines starved because of prioritization of higher request amounts.

What I mean by "simple" is they completely avoid "OK now I have to split off several belts of components from the bus, and figure out how to get all of them through here in a limited space".

Even with a "perfect" bus setup, you are limited in throughput by the amount of resources that blue belts can carry. A furnace collumn can't be more than 72 furnaces long because that's the maximum a blue belt can carry. Just adding more and more belts, you'll always eventually run out of room somewhere.

Indeed, but building a megabase is far different that using robots for a "regular" factory. In that case, I mostly have outposts that build intermediate products... including fully beaconed furnace setups that no longer have to be 36 machines long :-)

In fact, I would say that high throughput factories are better served by belts than bots because of the energy cost, speed, and carrying capacity of bots, unless you want to have tens of thousands of them.

My Belt Diva factory grew to launch a rocket every minute without the use of any logistics bots :-) I never ran out of space....

Originally posted by jchardin64:

Not necessarily. You can always put conditions on the inserters to only operate if the amount of material in the network is >X. Use different X for different things to control priority. For example:

--Inserter feeding iron into gear machines only operates if the amount of iron in the system is > 100. The output inserter operates if the amount of gears < 200.
--Inserter feeding iron into pipe machines operates if the amount of iron in the system is > 250. The output inserter operates if the amount of pipes < 400.
--Inserter feeding iron into steel smelters operates if the amount of steel in the system is < 1000 (and it doesn't care about iron supply).

So steel gets first priority until its quota is met. Gears have a higher priority than pipes do, so if there isn't enough iron coming in for everything then pipes will shut down first, followed by gears.

.... which kinda negates your arguement of being "simple" XD

[The Chaotic Coder]

Originally posted by KatherineOfSky:

Originally posted by jchardin64:

Not necessarily. You can always put conditions on the inserters to only operate if the amount of material in the network is >X. Use different X for different things to control priority. For example:

--Inserter feeding iron into gear machines only operates if the amount of iron in the system is > 100. The output inserter operates if the amount of gears < 200.
--Inserter feeding iron into pipe machines operates if the amount of iron in the system is > 250. The output inserter operates if the amount of pipes < 400.
--Inserter feeding iron into steel smelters operates if the amount of steel in the system is < 1000 (and it doesn't care about iron supply).

So steel gets first priority until its quota is met. Gears have a higher priority than pipes do, so if there isn't enough iron coming in for everything then pipes will shut down first, followed by gears.

.... which kinda negates your arguement of being "simple" XD

Once again:

Originally posted by jchardin64:

What I mean by "simple" is they completely avoid "OK now I have to split off several belts of components from the bus, and figure out how to get all of them through here in a limited space".

Even with a small factory that only supports 15-20 labs and a handful of machines that produce stuff like inserters and assemblers, no matter how well I plan ahead I always find myself having to move stuff around to make room for an extra belt of something to come in.

The control setup I was describing isn't as complicated as it sounds -- it only takes a moment to set a condition on an inserter. Then with blueprints you can paste it over and over again without having to do anything else. It may be a little harder than "just put a splitter on the bus" but it's simpler than "put 3 splitters on the bus, now figure out how to get the belts under each other so they all get to the production area, now braid the belts together to try to fit them between the machines and the beacons." It also gives a much finer degree of control over production, and eliminates unneccessary (and expensive) belt storage.

[KatherineOfSky]

@Jchardin -- I read what you wrote. No need to "once again".

Running out of room -- what you are describing is a lack of planning, not a deficiency with belts. You might try building on one side of the bus instead of both. I plan on 4 belts iron, 4 belts copper, 4 belts green circuits at the least. I then convert my green circuits to off-bus, requiring 4 belts copper and 3 belts of iron.

I don't move anything in my factory unless I am moving it to a completely separate outpost in end-game. I do plan, and a lot of that just comes with experience.

Also, once you learn how to split a line off the bus it is very easy. It's not a complicated math problem that requires a different solution each time. I would argue that it is a primary skill that should be learned because robots only arrive mid-game, and it takes a while before a new player would have sufficient quantities to ferry stuff around the factory regularly.

Also, even though it's isn't difficult, many people are quite intimidated by the circuit system if they haven't learned it yet.

[Lothos]

honestly, i find logistics bots lazy and "cheaty", even though i do use them.

I'm going to be rebuilding my current factory this weekend to be more elegant. Hopefully i can even reduce my bot usage a bit as well. Granted I'm doing good only using about 300 right now as it is.

[The Chaotic Coder]

Originally posted by KatherineOfSky:

@Jchardin -- I read what you wrote. No need to "once again".

I was trying to clarify (and doing it poorly). In response to OP's question, it IS much simpler to place a couple of logistics chests and let the bots carry stuff than it is to maintain a bus and split off multiple products every time you want to make something new.

In response you what you were saying about controling flow of materials to avoid overproduction in one part of the factory while another is starved, I gave a more complicated example of how it can be done, and how the logistic system actually allows much finer, more precise control of the machines. It's just a matter of changing the values in the inserters' conditions.

Anyway, I realize my "once again" may have been more harsh than intended and I appologize if you were hurt by it.

Originally posted by KatherineOfSky:

Running out of room -- what you are describing is a lack of planning, not a deficiency with belts. You might try building on one side of the bus instead of both. I plan on 4 belts iron, 4 belts copper, 4 belts green circuits at the least. I then convert my green circuits to off-bus, requiring 4 belts copper and 3 belts of iron.

Here's the disadvantage to that level of planning: Suppose I just started a game, and I'm building the beginings of the bus. First thing I make is gears (just 1 belt) and then green circuilts. You suggest I should leave room for 4 rows of green circuit machines feeding 4 belts, right? Well, I can easily leave myself that much space -- but that might push the whole rest of the factory so far that I run into the water before I've researched landfill. I might have to chop down a forest by hand. I might run into a biter nest when I don't have any bullet upgrades or power armor. At the very least, now all my belts have to be significantly longer, at a stage when I'm crafting most of them by hand.

Originally posted by KatherineOfSky:

Also, once you learn how to split a line off the bus it is very easy. It's not a complicated math problem that requires a different solution each time. I would argue that it is a primary skill that should be learned because robots only arrive mid-game, and it takes a while before a new player would have sufficient quantities to ferry stuff around the factory regularly.

I would suggest that once you learn how to connect requester chests to machines and set limits on inserters it's very easy as well, not a complicated math problem either. When I first started playing I thought that oil processing was incredibly confusing, but 1700 hours later I can set it up in my sleep.

In the end, there are things that belts do better and things that robots do better. It's a matter of personal preference -- but I use bots everywhere I can because my computer is something of a dinosaur. In January, I built a no-bot factory using trains and manufacturing outposts that could launch a rocket every 3 minutes, but by the end I was down to only 35-40 FPS. I'm trying it again using only bot, and while I haven't nearly gotten there yet I think it will do much better.