Factorio Forums

For me, the behavior or storage and active provider chests is the most annoying aspect of logistics right now. After reading several threads about it here and on reddit, I've come to the impression, that lots of other players agree on this regard.

Another proposal directly linked to the issue is here:
viewtopic.php?f=6&t=38540&p=230626

The problem should be well known: The current logic of bots makes them transport any actively provided or autotrashed items to very remote storage chests. This leads to bots flying across the whole base and funny-looking robot snakes, sometimes bouncing off full chests multiple times.

I understand that the devs do not want to add a filtering option similar to train cargos. I can see why, as this would improve the chests to a degree that should be the domain of circuit logic. Hence following idea: What about allowing a preferred or even exclusive content? This could be specified in the area where requester chests set their requests. You can still insert anything manually or by inserter. But when choosing the target for unloading any item, the bots should choose the closest preferred chest; with preferred meaning: Either contains already an item of that type or has this type explicitely set as preferred content. If you dont set any preferred content, you get the current behavior.

This behavior can be already achieved by priming each intended target chest with a single item of the target type. Of course, this is a horrible workflow and fails completely, if robots start to take items out (what usually happens at a train station, for instance).

With this idea, the most common use case is easily covered: At a train station for, lets say, iron plates, you unload into active provider chests and you place a large area of storage chests with preferred content iron nearby. As long as these storage is not full, you will not see bots travelling kilometers to a remote storage chest that contains a few iron plates. If your storage is overflowing, it's your fault, of course.

The advantages in my opinion are: This does not need large changes in bot logic, it can use existing GUI layout and perhaps even existing data structures, and it contains the current state as a default case.

I did this as a proof of concept:
https://youtu.be/YOVpZ9fTufk

With active provider and storage chests there's no control. Bots do whatever the devs programmed them to.

With passive provider and requester chests there's full control. Bots do exactly what the requester chests tell them to do.

The above works in a sequence like this:

• There's some items in some requester chests
• There's no items at all in any provider/storage chests
• Some stack inserters move some items from requester chests into passive provider chests
• Requester chests at destination are set to request those items
• Bots are dispatched to move items from provider to requester
• As soon as bots are dispatched the items are removed from logistic storage (even before actually being picked up)
• Once all required bots are dispatched (no items left in logistic storage) the requester chests are cleared from requesting
• Next set of stack inserters move some other items from some other requester chest into passive provider chests
• Requester chests at new destination are set to request those items
• Process repeats

This can be done by hand as well, but it's obviously way too slow and not automated. It's doable using circuit network though.

Process can be scaled to higher throughput via more chests+bots. Bots distribute items evenly among target requester chests. Logistic requests are filtered to begin with.

I think it might be possible to create a train unloading station using this method. Will require lots of tweaking though.

In my opinion all this issues and much more could be solved, if we would be able to have more than one logistic network.

Like this suggestion: viewtopic.php?f=67&t=8905 Overlapping Logistic Network II
A little bit long read and complex.

More ideas are here: viewtopic.php?f=80&t=18093 Roboport/Logistic Network/Robot enhancements
Look below "Logistic Network extension"

With that suggestion you can place the chests to store iron where you want, assign it to another logistic network and then create "exchange points", which fill or empty this storage just as you need. And much more.