Ant Colony AI

Overview:

Ant Colony AI is a DLL plugin for an ant arena program that uses an existing proprietary engine, the Avatar Engine. In this program, each ant colony DLL competes in a turn-based survival game mode where the goal is to be the last surviving colony in the arena. There are 4 different ant types (queens, workers, soldiers, and scouts) each with different abilities and actions that can be performed to benefit the colony. My DLL competed in a bracket-style tournament against 13 other programmers and got first place overall!

Tools: C++, Avatar Engine
Development Timeline: March 2024 – May 2024

Project Features:

Ant Abilities

The 4 types of ants in the program are the queens, workers, soldiers, and scouts. Queens are the only ants that can consume food and can only die once the colony’s nutrient count is at 0. Workers are ants that can pick up food and bring it back to the queen, dig up dirt tiles, and can create corpse bridges to allow other ants to cross water. Soldiers are the ants that can attack any ant type from another colony either eliminating the opposing ant or reducing the attacked colony’s nutrient count. Finally, scouts are ants that can traverse the map at less of a nutrient cost and have a greater sight radius to help the colony explore faster and decrease its fog of war. Below is a downloadable file for all the game rules of this program.

Ant Colony AI Rules Download

Map Types and AI Navigation Schemes

There are 7 separate maps each with different styles of terrain that the ants must be able to navigate around based on the current state of the map. Each ant uses distance fields and A* pathing to correctly navigate around any obstacles and to go to their desired targets. To the right are examples of all the different map types that can be used.