Could it be possible instead of using a Bruno method that horses are made into animals that the shepherd takes care of? So the horses would be hitched to a cart which could say carry 20 items and would be deployed only when you would drag out a zone. A worker would hop on the horse and ride next to the location, stop the horse, collect all the items, and ride back. It would remove the need for any “smart ai” in some ways.
I mean… theoretically, anything is possible.
In practice: you’d need to code two entirely new major systems – a “brain” for the horse AI, and a custom pathfinder add-on to get the hearthling to use the horses (you’d have to seriously upgrade the hauling logic too.) Ironically this method wouldn’t eliminate the need for a smarter AI, it would require a new AI which allows the hearthlings to “understand” what a horse is + how they work, and be able to plan a hauling route with the extra “go get a horse” step, AND understand that the horse has its own separate inventory which the hearthling needs to be putting things into and taking them out of.
The reason I suggested using cricket golems as a base is that their brain already includes hauling logic, so the “main” part is done for you – assuming that you’re ok with a My Little Pony-esque “horses can load carts by themselves” situation.
However, if you’re trying to add horses as a living entity which another entity interacts with (accessing inventory)… well, that’s a cool idea but there’s no precedent for it so you’d have to code it out yourself because there’s no “hook” in the game for that behaviour already. And like… yeah, someone who knows how to code their own game logic could probably figure it out; but at that point you’re starting to dig into the guts of the systems that the whole game relies on, so you start running into other concerns e.g. maintaining compatibility and making sure not to break existing systems.
The reason that nobody has made a successful vehicle mod yet is that it’s simply a whole other level of complexity on top of an already complex/expensive-to-process system and the sheer scale of work involved tends to be more than hobbyist modders can comfortably chew their way through. And, assuming someone does do all that work, you then boil down to the most basic effort/reward question: what did it achieve or add to the game? Like, horses are cool and it would be an amazing technical feat… but currently, we can just place down a vault and fill it up before moving it if we want to haul 256 items at once. Horses wouldn’t solve the “pathfinding for hauling tasks eats up the CPU’s memory” issue, and wouldn’t be significantly more effective than hearthlings or cricket golems unless you can ensure they’re picking up a lot of items in one place before they turn around and come back (at which point: go back to the above trick with containers for bulk-hauling, that actually works even better than a single hauler with a massive entity because it turns lots of long hauling trips into lots of very short ones + one large one at the end)
So, I’m not saying you shouldn’t try to make it happen; I’m just trying to point out the practicalities and pitfalls. Not because I don’t want to see it done… but because I don’t want to see it get halfway and then turn out to not be worth it, hahaha! But if you can come up with some way around those issues/limits, I’m keen to see it!