Discussing the (missing?) 4th type of transport!

[ssilk]

I come from this discussion in the suggestions-board: https://forums.factorio.com/forum/vie ... =6&t=13122
and created this thread, cause I really think, that there is a need in the forum to discuss about something how we want to handle long distances in general.

But that shouldn't be done in the suggestions board.

So I did it here.

Related subjects are also:
https://forums.factorio.com/forum/vie ... f=6&t=3072 Outpost logistics
https://forums.factorio.com/forum/vie ... f=6&t=2223 Factorio <> Industry Giant 2
https://forums.factorio.com/forum/vie ... f=6&t=7150 Pneumatic Delivery
https://forums.factorio.com/forum/vie ... f=6&t=1941 Long-range bots alternative (especially here: https://forums.factorio.com/forum/vie ... sts#p14046)
https://forums.factorio.com/forum/vie ... f=6&t=3802 Rocketry and Asteroid Mining
https://forums.factorio.com/forum/vie ... =6&t=13086 Orbital base construction support

So before we are going to discuss, I want to define

Exclusions
- We are NOT speaking about transport of the character
That is a completely different theme and there is also not that many side-effects; it will not break the game, if we introduce a teleporter for the player. That has already been implemented in several mods and that works fine. See also https://forums.factorio.com/forum/vie ... it=telepor* where Kovarex himself philosophized about Teleports for the character only and said relatively clearly, that a teleport will only be useful for character transport.
- We are NOT speaking about mass transport over long distances
Because fast mass transport over long distances is in my eyes perfectly handled by trains. There is also a need for slow mass transport over middle distances (something between belts and train, see ropeway for example: https://forums.factorio.com/forum/vie ... f=6&t=7977 ), but that is also off-topic here.
(Other stuff: https://forums.factorio.com/forum/vie ... f=6&t=2247 Chest-on-a-belt, https://forums.factorio.com/forum/vie ... f=6&t=1157 Roads, https://forums.factorio.com/forum/vie ... =67&t=4529 Large Bridge Cranes )

So what we are talking here about?
- Transport of items
- Over large distances (in general over 500 tiles, up to 10.000 or more)
- Small amount of items (hundrets, maybe 1000, eventually 5000 (one chest))
- Very, very fast (transport should not take longer than a minute, from request to delivery)
- Very fine control of what is transported. Because you cannot transport much items, you need to control the item types or the flow or whatever. This control is difficult and I don't want to mix this up with questions about how this connection is built with circuit network, because it doesn't make sense to connect an outpost by circuit network over 5.000 tiles (To built the connection you need currently minimum 1 hour - and this is needed for every outpost). An idea about this type of transport should also include how to transport this information.
I found a post about this: https://forums.factorio.com/forum/vie ... f=6&t=2108 Logistics storage fixes

And why?
- For building purposes in general: You are at an outpost and want to built a mining site for example. Instead of bringing everything in a car or in the own inventory and then find out, that we forgot the salt, we can use this.
- It is fun to drive by train or whatever, even if it takes several time, but it is just boring to drive such long distances over and over. It doesn't make sense.
- Speeding up the game. So that it makes more sense.
- For building with blueprints.
- For automated building.
- For more or less centralized building.
- For Multiplayer to deliver requests without walking to the other player (which might be impossible, if the spawns are far away). To enable game-modes, where not everybody has or can produce everybody.
- For game-types where you have tasks and need to deliver things very quickly, no matter about the costs.
- Things we don't know yet. I'm sure we find them.

What ideas/suggestions are already existing?
See linked list above!

No seriously: Ideas to avoid misusing it as mass-transport:
- Transport via rockets. A one-way transport, fill rocket, fire it, flies to target (like in the Long-range bots alternative)
- Using a transport-capsule (Cargobox) which cannot be transported empty (like in the pneumatic tubes idea)
- .... more ideas?

Ideas to request the items:
- Just have a control to fill a "chest". Like for the rocket-idea.
- Some kind of "finding out the needed items" in the target area and requesting it in the source area.
- .... more ideas?

To the last point this is related:
https://forums.factorio.com/forum/vie ... =67&t=8905 Overlapping Logistic Network II
I think it is a precondition, or side-condition to have better separation of logistic networks.

Erm. And what should be discussed?

Everything.
At first I think: Is this really needed? Where should the game go to? Do we really want to have Factorio more into the direction of OpenTTD/Industry Gigant?

[Klonan]

Something like truck stops and trucks would work well, with the truck taking a medium inventory size. At the truck stop, you use a logistics request slots to order what you want. After finding where these materials are in any logistics system, a truck can be dispatched to deliver the stacks of materials, from any other truck stop in all systems.

With the truck stop similar to a train station, the trucks are essential logistics bot that travel over ground, but carry a much larger amount of cargo, which travel faster than belts, but slower than trains.

[ssilk]

Hehe, trucks don't hit two needs:
- Not very fast (slower than trains)
- Not small amounts of items (many thousands, instead of many hundrets)

Eventually airplanes (like suggested in https://forums.factorio.com/forum/vie ... =10#p88230 ) hit this needs?

[Takezu]

Well trucks where a light cargo alternativ in close/mid range, but i don't really think for long range. They'd be to slow i think.
A truck would maybe came up to 100km/h great on short distances but the farther you go the more it reletivates.
It's a good idea none the less.

For really long distances, i'd go plane, same concept but different speed, on short distances not worth it due to starting/landing time, but on long distances,
it could really play it's speed trump out.
And with limited capacity, because after all for mass cargo the train is unbeatable. It's theoretical unlimited just ad more wagons and maybe loks.
The stops could act as requester in double sense, requestin stuff for transport and stuff getting transported, define a route and at the end of the route the logistics can request things it needs. The other points at the route could request the things from ther lokal Logistic network if it's provided.

I mean somthing like, my outpost needs 20 Repairpacks 15 turrets and 2 spare roboports. Point A provides Turrets, point B provides ports and Packs.
A Truck/Plane/whatever starts at point a loading 20 Packs, moves to point B loading 20Packs and is full (pretending only two slot size or an other limit in cargo?). A spare Transporter loads 2 Ports in point b and both make their way to the outpost. Unloading and waiting their for further transports.

Maybe i'm overshooting a bit here. But that would be the ideal?

[Koub]

I love the idea I've seen in a mod : rocket delivery. It's so totally badass
https://forums.factorio.com/forum/vie ... 32&t=12424

Transport of items
- Over large distances : Check, the distance limit could be any, there could even be tiers of research to upgrade rocket range.
- Small amount of items : Check, can't put many items on a rocket, but you can build several of them for quicker delivery.
- Very, very fast : Check, rockets are faster that any actual transportation system.
- Very fine control of what is transported : Check, in the mod, you can request whatever you need, as in a requester chest.

[vampiricdust]

I suggest the plane for two main reasons:

First, they're already placeholders in the game for them. I don't know what the devs intended for it to be small transport vehicle or just player transportation.

Second, planes won't have to worry about pathing. A truck would have to path around stuff or require a road of some form to guide it around, which makes it like a train too much. Anywhere at truck would be used, logistics bots would mostly be better due to the lack of pathing. I'm not saying pathing is bad, but it is one of the most computationally heavy task for a computer program to do. Planes would not have to worry about pathing.

Planes, what do I mean?

Planes would have very limited cargo space, maybe 2 or 4 slots. This would be enough to bring a couple stacks to you and can perform an air drop of the items (one way transportation). The other use I think would be cool would be maybe a cargo plane with say 10 stacks of space. A train is still infinitely better as you can put many more cargo wagons on to get the storage you need. The planes would travel between airports. These would be great for carrying supplies like turrets, repair packs, spare fuel, ammo, or any other odds and ends you might need, but don't want to dedicate a train to carrying around. Planes could fly on demand rather than a running a route.

[SHiRKiT]

I agree with @vampiricdust and @Koub

The downside of Rocket Delivery (as it's currently implemented) is that the items arrives when the rocket is launched, and not when the rocket arrives =\

And about planes, if routes are implemented, it would be pretty insane, since planes costs so much fuel, and having a plane on a route would mean so much fuel being used. I like this idea if Solid Fuel is used for this, or maybe implementing something crazy like the Rocket Delivery.

I love those concepts!

[Lupoviridae]

I like the idea of a transport ship/plane. I'm imagining something that looks a bit like a a quadcopter. You build a 3x3 Heliport at your main base and one at the destination location. It flies at a moderate speed back and forth, or maybe in a loop between multiple heliports carrying a mid-size item payload (1000 items or so). Heliports would be low material cost, but the drones themselves would be more akin to building a train locomotive. Basically the airport idea, but more factorio-esque. Heliports would be designated either supply ports or drop-off ports. Each supply port has 1 drone. Supply ports request directly from logistics. Resources can either be directed from the supply port (bring 1000 iron plates to port A), or requested from a drop off.

Basically, you tell the supply port at your main base to bring 1000 iron plates to outpost A. Logistics robots carry the plates to the heliport. Once full with the plates, it takes off vertically and flies linearly to outpost A, drops off the items, then returns to await another request.

[bobucles]

It's difficult to make fuel use "oppressive" because coal is easy to set aside and solid fuel is free. Because of this, planes eventually become a method of "trains but better". There is nothing stopping a player from building 10 or 50 or a thousand air pads to haul all their goods, doing away with the rail network entirely.

If your high tier transport burns copper and iron as upkeep, then it becomes VERY expensive to use VERY quickly. Such a system can't be used for regular transport because it will burn more raw goods than it produces. And if it's slower/weaker than trains, why not use trains?

The main reason a player needs arbitrary items in small quantity is to use them for construction. The need for items is immediate, and the player can be anywhere on a vast map. So why not just teleport them in? It can cover any arbitrary distance and still keep players connected to their home logistic network when they're thousands of tiles out. Make the process consume some high end raw materials (adv. circuits, proc units, steel), and it'll do exactly what it needs without being exploited for EVERY item.

[Takezu]

How do they say: Everything worth doing is worth overdoing?
If a player choose to massproduce an expensive, limited in cargo, method of transport, then thats on him.
But sooner or later the UPS will enevetably went down to a point where it's not funny anymore, because of the mass of entitys and pathfinding calls.

And on the other Hand, goods aren't transportet by burning iron and copper, that makes no sense whatsoever.

[Rockstar04]

How do they say: Everything worth doing is worth overdoing?
If a player choose to massproduce an expensive, limited in cargo, method of transport, then thats on him.
But sooner or later the UPS will enevetably went down to a point where it's not funny anymore, because of the mass of entitys and pathfinding calls.

And on the other Hand, goods aren't transportet by burning iron and copper, that makes no sense whatsoever.

Directly using iron and copper, no. I agree that dosent make sense, but making it like a blueprint, where to reset it you consume a circuit could be more logical.

If you had to "program" say an advanced circuit with that specific flights information, and then discard it to program a new delivery flight, that would make sure that this method of transport was adequately costly for the convenience it offers

[Balinor]

Trucks and truck stops make the most sense to me. For those saying that trucks would carry too much stuff then just limit the size of the truck to a few hundred items. Simples.

[Takezu]

Agreed, using an electronical chip for changing the destinations, or considert points on the route does make sense.
But, the cargo deliverd on the route should be something thats flexible handeled by the logistics.
After all it should be a longrange support for bases, so that repair packs and spare buildings can be ordered "from home".

woulden't make sense to me to nerf that into oblivion, or making that a manual step, because if the cargo handling where to be manual.
I could just throw the stuff in a wagon and drive myself, would be the same.
But true, setting up routes and/or changing them could cost, Advanced or even processing units. Makes sense to me.

Trucks are considert too slow
On short ranges they would be optimal i think, but reagrding ranges of 5K tiles and more, they just won't cut.

[Lupoviridae]

Trucks are considert too slow
On short ranges they would be optimal i think, but reagrding ranges of 5K tiles and more, they just won't cut.

This is why I like the idea of a quadcopter. It's like if you take a logistics robot and made it 100x the size. Completely integrated with logistics for loading. Slower and lower capacity than trains. Basically, instead of having 100 logic bots fly the goods out, one big bot gets loaded up and heads out.

@Bobcules It is true that someone could use this system to replace trains, but it wouldn't be nearly as fast nor as high capacity, so for most applications it doesn't really make sense.

[Koub]

This is why I like the idea of a quadcopter. It's like if you take a logistics robot and made it 100x the size. Completely integrated with logistics for loading. Slower and lower capacity than trains.

I thought the idea was, on the contrary, to be much faster than train, so that delivering a few items to a very remote outpost wouldn't take too long. That's why I like one-use rockets : you can balance it so that cost would be prohibituve for high throughput transportation, but usable (at high cost) for sending very fast a few items to a remote location.

[Takezu]

Well rather then a rocket, which is surly spectacular, i would have a "safer" transportation verhicle.
Maybe me but to me a rocket is a high speed projectile and delivering goods with it seems a bit risky.
Those things tend to explode.

A fast lightweight Aircraft would fit much better.

But yes, fast but limited.
Maybe with an overhaul of cargo. Something like defining a weight system rather then the working but simple Stacksystem.
With a weight system you could easily limit the Cargo. A stack system would allow always at least a stack, a weight sytsem would alow things like, only one of this heavy item but hunderets of an other light item.
Which doesn't necessarily would mean a drop of the current system, could easily coexist with it and only applied for Transporters like Cars/Trains/whatever.

[ssilk]

Takezu, why is safeness a need? This transport is not needed to be used for character transport. Therefore we can use real teleports. That is told so by Kovarex himself. See link above.

I think as Koub: The most important stuff is the speed. If we say, "We want to use this 4th transportation method mainly for construction", then the speed is here the most limiting factor!

Let's say we want to place a blueprint. Some mining sites, belts, poles, everything with blueprints. Blueprints place ghosts and ghosts have a timeout of 15 minutes. But 15 minutes is of course much too long to built something.

My idea about building by blueprints is that it should not take longer than one minute. One minute is of course much too short in most cases. Even if you built very near to your master-factory (no 4th transport) it can take minutes.

But one minute is a good target. Because if a transportation-method theoretically is not able to deliver something within this time, then it is not a good method.

So let's see, how this is theoretically working for transport methods like a plane:
- The time to bring the items to the plane (*), let's say in the best case 10 seconds, which is mainly the time a bot needs to come out of it's roboport fly to the source chest and bring it the plane.
- Then the start and flight of the airplane. Let's say it is 5000 tiles, which is far, but not that far. If a tile is one meter, that is 5 kilometers. It takes some time to accelerate and decelerate and after 500 tiles/second (1800 kilometer per hour, 1118 miles per hour), I think it begins to look a bit ridiculous. So 500 tiles/sec is the maximum.
So when you play a bit around with this page http://jumk.de/formeln/beschleunigung-zeit.shtml then you can see, that 20 seconds for accelleration and 20 for deceleration needs still 25 m/s² that it about 3 g.
- Then unloading and bringing the items to the target. I calculate that with 20 seconds, because the target is normally beside the landing zone.

So we are in sum at 70 seconds. And that is the theoretical maximum!

When I calculate that in about the same way for rockets, or for the pneumatic tube I come below this time of 1 min.


In other words: It makes no sense to put everything into some entity. Loading, transport and unloading takes for this kind of transport too long.


(*) We cannot do it like with the train and bring every item type we might need to the train station.

[Takezu]

Well who needs the one minute?
Who needs it really, there where great proposals for alternate low quantety transport.
For the gameplay it dosen'T really matter if it takes 60 seconds or 120 seconds.
Fast should be faster then now possible, not as fast as an arbitrary set value which is known to leave
nothing over then a few known from the beginning proposals.

And i can't see any gain from prefering an artifcial bend mechanic to fit, over an realisticly proposal, like the trucks, like the quadcopter
or like an aircraft, by simply declaring, it's ten seconds of my preset value it's not worth it.

And if we don't bring the stuff to the transportation vehicle, how should it be transportet?
Thats something i don't get, if i want to transport someting i have to get it into the transporter.
I have to load it in somway, that stuff won't fly over on its own.

[Klonan]

It sounds like the rocket idea is the best so far for long range and swift delivery, but it doesn't work if we time it from launch to landing.

What would make the rocket work in gameplay and logistic sense, would to have the rocket already up in space, maybe docked at some orbital supply depot. Wherever you want items, call down a rocket (maybe a droppod?) and theres your goods. After delivery the rocket can either destroy itself, or take off again back into space. Perhaps tiers of supply rocket could be used to carry greater and larger loads.

The rockets would need to be launched, deliberately, much prior to your request, back at base, and you would need to anticipate what you might need. sending up supply rockets with staks of miners, electric poles and inserters, and having the forsight to know what to send.

Supply rockets could be designated to automatically resupply the orbital station, when certain supplies are running low, leading to a space based rocket delivery network, and staying within the realms of a realistic transport system.

[Koub]

The thing is that 5000 tiles is very arbitrary too. How far are the furthest outposts in a Factorio game ? Probably quite further. the rocket is an excellent balance between speed, limited carrying cargo, and useability. For planes, you need landing strips, which is not that convenient.