NightAtMuseum

Night at the Museum | Gameplay Systems
← Back to Portfolio

Night at the Museum

First-Person Stealth Game | Unreal Engine | C++

A systems-driven stealth experience focused on AI perception, player visibility, and dynamic alert state transitions. Designed with modular architecture to allow scalable stealth behaviors and performance-conscious gameplay systems.

Unreal Engine C++ AI Perception State Machines Optimization

Project Overview

Night at the Museum is a first-person stealth prototype centered around systemic AI detection and tension-driven gameplay. The core objective was to build a scalable stealth framework that could support patrol systems, alert escalation, and player feedback loops.

Gameplay Systems Implemented

  • AI Perception-based detection system (sight + awareness thresholds)
  • Multi-state alert system (Idle, Suspicious, Alerted, Searching)
  • Player visibility evaluation logic
  • Modular patrol route behavior
  • Optimized tick logic to reduce unnecessary processing

Technical Challenges

One of the primary challenges was ensuring smooth transitions between AI states without creating erratic behavior. This was solved by implementing controlled state machine logic and clearly defined transition conditions.

Another focus area was performance optimization. AI logic was structured to minimize unnecessary runtime checks, ensuring stable performance with multiple active enemies.

What I Would Improve

  • Expanded sensory system (sound propagation simulation)
  • Dynamic difficulty scaling based on player stealth success
  • Tool-based stealth gadgets with modular interaction systems