When do you use logi bots?

[Jupiter]

I don't know if there is already a topic on this but I couldn't find it.

What I would like to know is when it is appropriate (useful/efficient) to use logistic bots instead of belts. I know that a common usecase is in large end-game-style smelter arrays with electric furnaces and beacons and all that but there should be more such usecases.
Using logi bots probably isn't always necessary as it is a relatively expensive solution, especially for low-throughput situations like the production of capsules or engines or other stuff you don't need a lot of. Bots and ports cost a lot of resources to make and they eat power.

I would love to hear where you use your logi bots for!

[garath]

I mainly use logistic robots to bring me stuff I need to use while expanding the factory including:

1. Transport Belt
2. Inserters
3. Underground Belt
4. Power Poles
etcetera.

I find it incredibly, incredibly useful to not have to run around picking these things up while I am expanding the factory.

[bobucles]

The first easily accessible use for bots is to repair and rebuild your base using construction bots. It only takes a handful to really ease the burden of keeping your base intact.

You don't really NEED logistics bots to get the rocket. However they are very good for fixing belt spaghetti. All it takes is a blue chest to bring materials into your assemblers, and red chests to move it out. This makes it really easy to build ammo and capsules, and even helps to build more bots.

[kingarthur]

ditto on the amazon drone deliverers. i also like to use them for items that only get used by the player in low volume like pumps because my electric engine usually get built near my oil setup and i never need more than a stack. so i dont find the effort to belt them worth it. also would start using them in mass when 2-3 blue inserters stops being fast enough to keep a belt feed assembler running full speed but the was pre .13

[Jupiter]

Thanks for the replies guys. I hope there are more coming in

My own considerations so far:
I would love if they bring me goodies all the time but, lets face it, it is a mere luxury. I'd rather setup a system that distributes the items around my base to various shopping points where I can pick them up whenever I want. It's much cheaper and the logistical challenge might even be fun.
Setting up the production of something new (ammo, miners or w/e) is very easy with logi bots but if you have a proper bus system then it is almost as easy to set it up with belt, and it's cheaper.

The late game smelting arrays is (probably, never gotten to that point) a genuine use case in which the bots really do make a difference so if your gonna build bots anyway then the above 2 points aren't relevant anymore of course.
But I also want to know if there are more situation in which logi bots really do make a difference.

[Miravlix]

I completely disagree with the ops reasons for not using bots.

Low throughput is ALWAYS something that you should use bots for, because any item on a belt is useless, unless the belt is moving that item at optimal throughput.

High throughput stuff like ore/iron plates is, because they get setup before you can make bots, a fairly good choice for belts, but bot based designs seems to not only lag less, but also be better for high throughput.

Bottom line is that bot is simply always better, but makes your "factorio" look like crap.

[Jupiter]

I completely disagree with the ops reasons for not using bots.

Low throughput is ALWAYS something that you should use bots for, because any item on a belt is useless, unless the belt is moving that item at optimal throughput.

High throughput stuff like ore/iron plates is, because they get setup before you can make bots, a fairly good choice for belts, but bot based designs seems to not only lag less, but also be better for high throughput.

Bottom line is that bot is simply always better, but makes your "factorio" look like crap.

If I'm reading your comment right, what I meant by 'low-throughput situations' are situations in which you don't need high throughput. Not that you're suffering from the lack of throughput.

[Kelderek]

If I'm reading your comment right, what I meant by 'low-throughput situations' are situations in which you don't need high throughput. Not that you're suffering from the lack of throughput.

Good examples of low throughput or limited need items are all the items you only ever need 1 or 2 stacks of at any given moment:
assemblers
furnaces
pumpjacks
chemical labs
refineries
chests
power poles
etc.

I like to use belts if I can put multiple assemblers along it, but I hate running a belt that only one assembler will use - that's a good opportunity to use requester chests instead.

[Qon]

When building a megabase, bots are the answer.
Bots are better for low and high throughput.
When I've made the best possible factory with bots I'll try making a belt based megafactory.

Bots > belts.
Bots <3

[MrGrim]

My current biggest use for them involves the mod Hard Crafting. It use them to sort the output from miners. The mod adds multiple outputs to mining ore. I load the mess unsorted onto a train, and that train goes to what I call "central sorting" which consists of several stops in a row each with 5 dedicated roboports with nearly 1500 bots in that one isolated network. Next to the unsorted ore drop off is a series sorted ore loading stations where different ratios of each output get loaded onto a variety of trains headed for processing facilities and waste management facilities. A few combinators controlling requester chests make sure proper quantities are delivered to the various outputs.

For vanilla, I will use them strategically. Maybe to fill a fuel chest for a train, for some low use items not worth dealing with belt spaghetti over, barrel management, etc.

[OBAMA MCLAMA]

The first use of them comes to when you get and make them. Either to make side items easily or to have your bots always fill you up with some iron plates. Both are usually the first use.
Such as making a machine on the side that requests items to make a red long inserter, something you might not need a constant supply that you can just have auto crafted till later.
You will end up making more, almost one for every item maybe.

Eventually you make the decision to have a bot focused base, or a belt focused base.
Making a bot base.... is simple. Your bot network should hold thousands of bots, tens of thousands. And they move every resource to specific designs all over the place. Ore to furnace. Plate to circuits... etc.

And thats where it ends.

[Šedý]

I hate wasting resources. Coal, steel, electronic circuits... I need a TON of that, so it is fine to make a belt bus for them and run it through the whole factory. I don't mind these things lying on belts, because I will eventually use them all. But having a belt for processing units or modules? That's a LOT of unnecessarily buffered resources. I'd rather have a "standard robot assemby unit" (SRAU) there, instead. Requester >>> machine >>> provider.

Also, I think that nobody mentioned the speed of factory expansion. You can put blueprinted SRAUs anywhere in the factory. With belts, you have to run to the place where you have them. Of course, it is cool to have a specialized refinery, copperworks, engine factory, etc. It looks beautiful. But sometimes you want to expand fast... in which case robots are the way to go.

For example, I do whole rocket construction by robots, just spawning SRAUs everywhere around the rocket silo. When I go for the rocket, I have nothing else to do and I want to finish the planet as fast as possible.

[Jupiter]

I hate wasting resources. Coal, steel, electronic circuits... I need a TON of that, so it is fine to make a belt bus for them and run it through the whole factory. I don't mind these things lying on belts, because I will eventually use them all. But having a belt for processing units or modules? That's a LOT of unnecessarily buffered resources.

The point of bussing intermediate products is to compress your resources. If you make processing units early on and put them on the bus then a single belt of them can transport a whole lot more of raw resources than existing belts ever could. This reduces the need for 4 or even more belts of iron, steel and copper plates.

[MrGrim]

The point of bussing intermediate products is to compress your resources. If you make processing units early on and put them on the bus then a single belt of them can transport a whole lot more of raw resources than existing belts ever could. This reduces the need for 4 or even more belts of iron, steel and copper plates.

I'm not sure I understood you, but it sounds like you're describing buffering. If your supply can't meet demand bursts and if your demand is bursty to begin with then buffering between demand spikes can help so long as the spike isn't too long, but buffering can be done with belts or bots.

[Jupiter]

The point of bussing intermediate products is to compress your resources. If you make processing units early on and put them on the bus then a single belt of them can transport a whole lot more of raw resources than existing belts ever could. This reduces the need for 4 or even more belts of iron, steel and copper plates.

I'm not sure I understood you, but it sounds like you're describing buffering. If your supply can't meet demand bursts and if your demand is bursty to begin with then buffering between demand spikes can help so long as the spike isn't too long, but buffering can be done with belts or bots.

I was talking about throughput of your bus. You can either put a single processing unit on a belt or all of its ingredients separately. If you put the ingredients on a belt you might have to increase the number of belts.