Back to Projects
GAMESIN-PROGRESS

Quidditch AI Flight System

Behavior tree-driven AI agents that collect items, mount brooms, and navigate 3D arenas.

SOLO DEVELOPER  ·  Feb 1, 2026

Quidditch AI Flight System

Capstone project featuring a complete Quidditch AI flight system built in Unreal Engine 5.4. AI agents use behavior trees with Blackboard data to autonomously collect items, mount brooms, navigate to staging zones, and engage in aerial gameplay. The system demonstrates advanced navigation mesh integration, dynamic obstacle avoidance, and state machine transitions between ground and flight modes.

// TECH STACK

LANGC++ENGINEUnreal Engine 5.4SYSTEMBehavior TreesSYSTEMAIBlackboard

Contributions

  • 01Designed and implemented behavior tree architecture for AI flight navigation
  • 02Built custom Blackboard keys for dynamic decision-making during gameplay
  • 03Integrated NavMesh pathfinding with custom flight movement components
  • 04Created staging zone system with trigger volumes and state transitions

Technical Highlights

  • ▸Behavior Tree with 12+ task nodes and 4 decorator conditions
  • ▸Custom EQS (Environment Query System) for item detection
  • ▸Smooth ground-to-flight transition using interpolated movement modes
View Source