Approach to make A.I similar to in Two Point Hospital?

Hi guys, I'm currently making kind of city building game and i want to make some NPC behaviour that interact with something in the game like in Two Point Hospital. I'm not trying to make it as advanced as Two Point Hospital. Their NPC has some parameter like hunger and thirst that decide it's action. I only need basic things like NPC comes in, do thing, and maybe do another thing, and get out.

I'm currently thinking of using NavMesh Agent and Add some commands and queue their action when they spawn. Why NavMesh ? because there will be multiple level of it. I thought using Paths will need more time since every new level i need to add more paths.

I would like to know your approach to make something like this and try to make efficient as possible since i'm targeting mobile platform

Here is trailer of Two Point Hospital in case you didn't know it

 

You can use the concept of Utility AI to let each NPC decide what to focus on based on a variety of different needs.
https://en.wikipedia.org/wiki/Utility_system

 

Agree. I'm currently researching utility AI for my turn-based tactical game. I'm reading the book Behavioral Mathematics for Game AI by Dave Mark. It feels like it's the one piece I was missing. You have all these AI systems like pathfinding, influence maps, behavior trees and state machines, etc, yet the single most important piece of the AI is the decision making, and Dave Mark's book really focuses on that, specifically for determining utility via score functions.

@rerwandi I'd like to point out that while you certainly need pathfinding, and NavMesh can help with that, it won't help with deciding hunger vs thirst. As @ShilohGames mentioned, utility AI will help with this, and the book I mentioned goes into practical detail on how to build such a system that can interpret utility and make a decision for your AI agent.

 

Looks like that's the thing i'm missing too. After reading some articles about it, i'm sure where and what i should trying to do. Thank you