University research project: Codename MGS

AI research applied to games


This customizable stealth game was created as part of a Master Degree’s final project. In it the enemies use several artificial intelligence techniques such as Ant Colony Optimization, Decision Trees and Neuronal Training.

Enemies have different characteristics such as visual and auditory ability, tendency to collaborate with their partners or tenacity, which make each one unique.

Other enemies such as dogs can use their nose to follow the player (based on Ant Colony Optimization) and alert other enemies.

The player on the other hand has several movement options, being able to crouch to reduce his visibility and noise generation or running to move quickly between points but risking being detected or getting tired.

In addition, the levels can be interconnected unlimitedly and even recursively.

The project was qualified with Honors by the University of Cantabria.