Behaviour Tree

My BT is set up around 2 main flows:

1. Patrol flow (default):

   • Pick / move to patrol locations.
   • Keep looping while the player is not detected.

2. Reaction flow (after perception trigger):

   • Switch from patrol to reaction/chase behavior when the player is sensed.
   • If the player is lost, leave reaction mode and return to patrol.
3. • Shoot player when in range.

Blackboard keys used

• TargetActor - stores the detected player actor (or None when not detected).
• EnemyClose - (bool) - true when player is within range.
• CanSeePlayer (bool) - true when the perception sense is currently detecting the player.

AIPerception setup (exactly one sense)

I used Sight as the only active AIPerception sense.

It creates two clearly different states:

• Before detection: AI is patrolling.
• After detection: AI visibly reacts (switches into chase/alert behavior).

When sight is lost (player hides/leaves view), the AI clears detection state and goes back to patrol.

This repository includes Git LFS tracking

• Patrol before detection
• The sight trigger causing a visible state change
• Attack/shoot when player is in radius
• Return to patrol after detection is lost

Video: https://youtu.be/mksy30LK_Is