Robots vs. belts

[Pascali]

hi, whats the better way to transport the things. Robots or belts? So is there need for belts in the lategame?

[BlakeMW]

Bots for short-medium range and trains for long range is the ultimate logistics strategy.

The main advantage of belts is they are really cheap compared with bots, especially the bot upgrades and the bot charging requirements can be considered to be significant expenses. But if you're willing to pay the price for bots they're basically better.
Don't use bots for medium-long range bulk transportation - if the bots usually have to charge more than once to make a delivery it's probably too far. Belts can be good for these intermediate distances - but so can trains. Good organization almost eliminates the need for such trips.
Beware that bots don't handle multiple sources and multiple destinations within the same logistic network very gracefully, for example if you have a big factory, and on opposite sides of the factory you have iron smelters, and iron ore is unloaded at both smelters and you want the ore to be delivered by bots from the unloading stations to the nearby furnaces, well some bots will be dumb enough to fly clear across the factory to get iron ore from the far unloading station. Bots are exceptional when you organize things well - using separate logistic networks connected by trains when necessary. If your factory is really disorganized and stuff is happening all over the place requiring frequent long deliveries with many charging stops, bots can be pretty bad.

[Zavian]

For a normal (ie not mega-base) factories build what you want. Bots make for simpler designs, but they also remove much of the challenge in factory layout. (They do introduce some of their own challenges, but they are normally easily solved for small bot networks)

Also the advantages of bots over belts in large factories is that bots are more cpu efficient compared to belts, which makes it easier to scale up to mega-base scale..

[Pascali]

Bots make for simpler designs, but they also remove much of the challenge in factory layout.

That is, was i think. I love the game, because of the beltsystem. That is 85% of the fun. And now i am getting to bots and they are more effective, if you are using some independent(some for turret-repair, some for fabriklane1, some for fabriklane2). So it is a bad feeling to have mor effecitve system but the system is less funny. In my eyes you have to start with bots and they have to get ineffecitve, if you build a good beltsystem.

I hope their won´t ever bee a possitiblity to define botsystems(with colors or something). In the end there will be a possibility to beam all things around. Be i am alone with this feelings? Factorio until gettings bots -> nice game! After that, getting a lot of more boring.

[Jap2.0]

Bots make for simpler designs, but they also remove much of the challenge in factory layout.

That is, was i think. I love the game, because of the beltsystem. That is 85% of the fun. And now i am getting to bots and they are more effective, if you are using some independent(some for turret-repair, some for fabriklane1, some for fabriklane2). So it is a bad feeling to have mor effecitve system but the system is less funny. In my eyes you have to start with bots and they have to get ineffecitve, if you build a good beltsystem.

I hope their won´t ever bee a possitiblity to define botsystems(with colors or something). In the end there will be a possibility to beam all things around. Be i am alone with this feelings? Factorio until gettings bots -> nice game! After that, getting a lot of more boring.

You don't have to use bots .

[Pascali]

Yes, but i know that they are more effecitve, so i wan´t. And the biters are going stronger, too. The repair-feature is nice to have too. The game would feel nicer if the transport feature wouldn´t be such a killer-feature. Maybe useful(working) for 2 (new) goods. But if you whole belt-network and system is shit, after robot possibility - it´s a bad feeling. A lot of time, i would build a nice belt, i put in 1 robot thing - done. Not the fun like with doing it with belts.

[BlakeMW]

There are other challenges other than belt layout. Bots empower beaconized setups (most typically alternating rows of speed beacons w/ productivity assemblers) that have tremendous assembler throughput, to the point where you might need 5 fully upgraded stack inserters to keep up with the assembler throughput - belts can't keep up with this kind of throughput (it works out something like almost an entire blue belt for one assembler), bots can keep up, but need very solid charging infrastructure, and then you have to somehow feed the voracious appetite of the beaconized setup, meaning complex train networks - and you want to pack in the train stations and roboports as tightly as possible to keep bot travel times down, but they don't fit neatly within beaconized setups - meaning more layout challenges.

Really, once you start really exploiting the power of bots (not just using them as a convenience) the primary logistical challenge moves to trains.

[Pascali]

hm o.k., i will see. But still i would like to have some fabric-setups where belts are the best solution.

[Zavian]

Well you can do a beaconed build with belts. It's just more challenging to design and layout and scale up than an equivalent bot based one, and uses more cpu cycles for the same output. (The highest throughput factories tend to be bot based, but that is probably because bot based factories are both easier to build and more ups friendly).

[Jap2.0]

It pretty much boils down to personal preferednce, and which you find more fun. Don't stress over it.

[Pascali]

if you play multiplayer and look at your neighbour, you wan´t to be nice too. So everybody would choose robots because it´s more simple with them. You can´t show your nice beltsystem... because no fabrication-lanes need this anymore. I would love to see, that 50% of the fabrication-lines would need or run much better with belts. Bots in factorio is like playing an egoshooter with autoaimbot.