
Define an npc state enum with passive, investigating, seeking, attacking, and dead, and wire it through a blueprint interface to trigger state changes and drive the animation blueprint.
Create a test bed level, set up a perception component, configure game mode and a first-person character, and implement basic damage and death for AI.
Retarget Epic's starter pack to a custom character, set up iron sights and aim offset, and integrate animations via the animation blueprint and montages for responsive Unreal Engine combat.
Explore perceptive AI in Unreal Engine using behavior trees and blackboards to sense players through sight and hearing, react with alert or attack modes, and patrol or idle when unaware.
Create and link base AI classes, including a behavior tree and blackboard. Initialize the guard AI controller with a combat state enum and navigation variables.
Develop modular state functions for air perception in Unreal Engine, defining combat state variants—attacking, investigating, seeking, and passive—and notify the Black Bull component of state changes.
Learn to build an idle system with a behavior tree in Unreal Engine, using selectors and sequences, decorators and blackboard checks to drive patrols, idle behavior, and combat state handling.
Build a patrol system in Unreal Engine using a vector point array, 3D widgets, and a blueprint-based move to task with wait times, random deviation, and looping patrol routes.
Implement a perceptive AI in Unreal Engine by wiring behavior trees, blackboard keys, and AI perception components to detect the player and drive enemy reaction.
Create an AI in Unreal Engine using a behavior tree and blackboard, implementing combat states, patrol logic, and a half-second timer service to switch between melee, ranged, and out-of-range behaviors.
Learn to mock up an ability system and interrupt handling in Unreal Engine blueprints, including an ability base, cast time, cool downs, and events for started, ended, and interrupted.
Explore how Unreal Engine AI uses behavior trees and a nearest-target approach to drive combat, movement, and targeting within the gameplay abilities framework.
In this series I will be showing how to set up two different AI designs, one for a shooter experience, and one for a RPG style AI. We will cover blackboard and behaviour trees, EQS queries, and the AI perception component using sight, sound and damage stimuli.
The Shooter example will be a longer, slower example where we will explain everything as we go, and it will use sight and sound to give it elements of stealth AI, along with a working shooting system driven by the AI controller and it's behaviour tree.
The Magic AI example will be a more concise, straight to the point demonstration of a different way of working compared to the shooter example, and we will utilise sight and damage stimuli to make a reactionary AI who is able to heal, cast spells, and teleport away from the user using a EQS query. This series also includes a large overview of the EQS system and how to many of the tests work.