When using logistic bots to fulfil logistic requests, without any cargo stack size research -> bots carry items 1 by 1
When using logistic bots to fulfil logistic requests, with cargo stack size research -> bots carry multiple items in the same trip
When using "ghost request" slots to send items to an inventory with construction bots, without any cargo stack size research -> bots carry items 1 by 1
When using "ghost request" slots to send items to an inventory with construction bots, with cargo stack size research -> bots carry multiple items in the same trip
When building the factory with construction bots, without any cargo stack size research -> bots carry items 1 by 1
When building the factory with construction bots, with cargo stack size research -> bots still carry items 1 by 1 <------------- this one is the inconsistency
Since 2.0, as I understand it, one bot can have multiple jobs assigned to it at once, so it already knows what it's going to do next after it finishes its current job. It should be that the bot is smart enough to realise (if there are multiple of the same entity ghost within its roboport network) that it is available to be assigned more jobs that are of the same entity type, and then do both in the same trip.
[2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
It’s simply how construction robots work.
If you want to get ahold of me I'm almost always on Discord.
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
@Rseding91 Then what is the purpose of worker robot stack size research affecting the cargo capacity of construction bots? Their main use is for building, not for transferring items between boxes - that's what logistics bots exist for.
Anyway, if I were to raise this as a feature request, is there any chance of it being considered? Consider how satisfying it would be to see construction bots going out and building 4 belts each when stamping belt lines, and how it would be a nice way to have charging speed be less of a bottleneck when stamping very large builds at once.
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
Deconstructing inventories and placing tiles.Zaspar wrote: Sun Jun 28, 2026 9:14 pm Then what is the purpose of worker robot stack size research affecting the cargo capacity of construction bots?
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
Oh, it works for placing tiles? Interesting
If you don't mind me asking, is the reason that it doesn't work for entities because of a limitation in implementing that, or an intentional choice? If it's a limitation, how did you make it work for tiles, and why can't the same logic be applied to entities? If it's a choice, why did you make that choice?
Thank you
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
Each robot is dispatched for only one job at a time. The scheduler does not plan for a robot to stop and build 4 different pipe ghosts in one trip.Zaspar wrote: Sun Jun 28, 2026 9:24 pmOh, it works for placing tiles? Interesting
If you don't mind me asking, is the reason that it doesn't work for entities because of a limitation in implementing that, or an intentional choice? If it's a limitation, how did you make it work for tiles, and why can't the same logic be applied to entities? If it's a choice, why did you make that choice?
Thank you
Building one entity requires usually one item, but can be more than one item like curved and diagonal rails.
Deconstructing one tree yields e.g. 4 wood.
Deconstructing one chest that contains 100 steel can fill each robot to the max until the contents are empty. Larger worker size speeds this up a lot.
Building tiles is the only job where one robot has multiple target ghosts when they all require the same tile. I suspect they need to be adjacent or close together to be grouped into one robot job.
My mods: Multiple Unit Train Control, RGB Pipes, Shipping Containers, Rocket Log, Smart Artillery Wagons.
Maintainer of Auto Deconstruct, Cargo Ships, Vehicle Wagon, Honk, Shortwave.
Maintainer of Auto Deconstruct, Cargo Ships, Vehicle Wagon, Honk, Shortwave.
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
If that is true, I see no technical reason why a bot couldn't build multiple adjacent entities in one trip. I ask again, why is it not the intended behaviour?robot256 wrote: Sun Jun 28, 2026 10:02 pm
Building tiles is the only job where one robot has multiple target ghosts when they all require the same tile. I suspect they need to be adjacent or close together to be grouped into one robot job.
Re: [2.1.8] Worker robot cargo size technology does not get utilised by construction bots when building entity ghosts
Zaspar wrote: Sun Jun 28, 2026 10:08 pm If that is true, I see no technical reason why a bot couldn't build multiple adjacent entities in one trip. I ask again, why is it not the intended behaviour?
The current behaviour is “Not a bug” because the game engine is doing what it was programmed to: assigning 1 Construction Robot to perform 1 Build job. If an entity requires multiple Items to construct (eg, Curved Rails) then the Bot will dutifully use the extra cargo space provided by the bonus to do so. Tile placement can involve a lot of entities at once so there is an optimization for Robots to try to complete multiple Jobs using the same Item type.
Asking for new behaviour that a Construction Robot should attempt to build multiple Ghost Entities of different Types (crossing multiple Jobs) during the same Flight would go in the Ideas and Suggestions forums. This would likely add complexity to the dispatching function, since these Items would not be part of the same Stack, and that is most likely why no Developer has even tried it yet.
Good Luck!



