Is there a name for sub-field of research focused on achieving some goal (possibly computational) with human and/or other imperfect agents?

One is probably solving for some real physical world "UI/UX" considerations on one hand, possibly in dispersed/async conditions, and robustly managing unreliable agents on the other hand.

I suppose the highest current form might be the augmented humans in amazon warehouses. But what was the best coolest way to run a warehouse before any high-tech augmentation was available?

Also that thing about how UPS drivers only make right turns? Traveling salesman with a team of competing salesman in one market? Best way to unload a truck into specific rooms of a house? Without running a python script on a laptop because that's not a real solution anyway?

I would imagine there is a rich history and retrospective research on how pre-telegraph societies implemented particularly effective and clever dispersed problem solving. Maybe draw a line somewhere below the huge scale/complexity of command-and-control military theory or whatever. That line might be that the research aims for and achieves discrete prediction of success and the time/energy spent. Maybe just a corner of operations research?

Or cool histories of whatever I'm talking about!!