Surviving State Tree: A real case study with Dinosaurs
summary
Gonner uses Unreal Engine 5 State Tree to replace brittle behavior trees and ships a working pipeline: multiple navigation grids via Maruna navigation plugin for creatures from chicken-size Ricardo to a ~20 m Alamosaurus, a perception stimulus sorter that drops stimuli older than 5s and scores by team, threat level (examples: threat 5 vs 2 vs 8–9), and senses, then a state response selector that produces prioritized transition requests. A custom toolbox of move/rotate/animation, Gameplay Ability, EQS and meta tasks (run StateTree, send event, block transition) speeds development and stabilizes AI.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:32Project Design Requirements
- 3:28State Tree Overview
- 8:33The Solution
- 12:23Pipeline
- 17:47Development
related talks
59mGame AI 4N00bz
35mHero AI – Gameplay Abilities Meet Behavior Trees
52mThe Journey of AI in Games: A Technical and Philosophical overview
38mPractical AI in Unity
01h 01mThe Forgotten AI of Video Games
43mOptimizing animations in Unreal Engine 4
40mModulating Game Difficulty with Machine Learning
39mEmbarrassing Mistakes: a short story about AI