Search this site
Embedded Files
Derek Simpson
  • Home
  • About Me
  • Projects
    • Well Witches
    • Malisense
    • The Patrol
    • Pumpkin Run
    • Chess AI
    • LLM Rythym Game
    • Deep Sea Save
  • Extra Curriculars
  • Research
Derek Simpson
  • Home
  • About Me
  • Projects
    • Well Witches
    • Malisense
    • The Patrol
    • Pumpkin Run
    • Chess AI
    • LLM Rythym Game
    • Deep Sea Save
  • Extra Curriculars
  • Research
  • More
    • Home
    • About Me
    • Projects
      • Well Witches
      • Malisense
      • The Patrol
      • Pumpkin Run
      • Chess AI
      • LLM Rythym Game
      • Deep Sea Save
    • Extra Curriculars
    • Research

Play The Game

Enemy AI Programmer

As a Programmer, I played a key role in developing AI systems, focusing on creating modular and adaptable frameworks. I implemented state machines to drive various enemy behaviors, ensuring that the AI could handle a wide range of interactions in a dynamic and responsive manner. My work on the core AI systems for Malisense emphasized ease of modification, allowing for seamless updates and scalability. Collaborating closely with designers, I helped shape the behaviors of all NPCs, ensuring the AI aligned with the game's vision and enhanced the overall player experience. 

My Work

Patrolling and pathfinding

  • Developed a designer-friendly patrol system for easy customization of enemy movement.

  • Enabled designers to define patrol paths by selecting areas, allowing the AI to navigate to random points within each selected rectangle in sequence.

  • Created a shared patrol system for all enemies, streamlining the design process and ensuring consistent behavior across different AI types.

  • Integrated the A* Pathfinding Project Unity library for precise and efficient pathfinding, enabling enemies to intelligently navigate complex environments.

  • Enhanced gameplay by providing adaptable and efficient patrol routes for all monsters, improving overall player experience.

State Machine

Each enemy uses a custom state machine with three core states: "Patrolling," "Alerted," and "Chasing."

The Sound Beast reacts dynamically to player-generated sounds:

  • Transitions from "Patrolling" to "Alerted" upon detecting a sound, scanning the environment for its source.

  • Enters the "Chasing" state if the player makes another sound within a defined radius, actively pursuing the player.

This layered behavior enhances AI depth, resulting in more immersive and unpredictable enemy interactions.

Perception

Monsters in the game are designed around four key senses: touch, sight, hearing, and smell, each contributing to unique behaviors and player interactions. The Sight Beast uses a vision cone, implemented through vectors and dot products, to detect players within its line of sight. The Hearing Beast detects sounds made by the player within a specific radius, triggering alerts and initiating chases when noise is heard. The Smell Beast tracks players through scent trails, allowing for strategic evasion as players must outmaneuver its keen sense of smell to stay hidden. Meanwhile, the touch beast acts a landmine if players touch it. These sensory mechanics create varied challenges for players to overcome.


GitHub Repository

Google Sites
Report abuse
Page details
Page updated
Google Sites
Report abuse