Omar Bazzi Game Dev
Greybox Drone & Guard AI

Greybox Drone & Guard AI

Prototype Unreal stealth AI

A few months ago, a video from a couple of years ago, from the official Unreal channel, appeared as recommended in my YouTube feed, where they reviewed the behavior tree system using a drone. I thought it would be a good opportunity to do something new (a flying object using NavMesh and BT) as well as to review some AI concepts in Unreal. I also included some guards with path following (something similar to what I did in this game a few years ago with the monsters, only now I added the sense of hearing and some extra behavior).

The video on the Unreal channel can be found at this link. The pertinent warning is that this is not a step-by-step tutorial, there are many things that it does not show how to do and for which you need to have previous knowledge (for example of Enumerators or a broad handling of Blueprints).


A summary of the most relevant applied contents:

  • Behavior Tree (Services, Tasks, Decorators), Nav System - For guard AI and part of Drone AI.
  • EQS (Generator, Contexts, Tests, Testing Pawns) - For searching for Drone patrol points.
  • Anim notifiers - For connection between character steps and enemy hearing sense.
  • Mixers, Sound Cue, Metasounds - For sounds of character and guard steps, and various Drone sounds.

Complete project (project repository in Azure DevOps)

Video on YouTube: Video on YouTube


Learning sources:

Unreal Engine AI with Behavior Trees | Unreal Engine

Stealth AI | Ryan Laley