Artificial Intelligence Projects
Goal Oriented Behavior
This is an AI meant to keep itself alive once input has been given based off of a discontentment tracker. Once input is given it will try to resolve discontentment in hunger thirst and exhaustion in the most efficient way possible. I have modified the discontentment values to make sure all possible actions are likely to be used and I created a text display for the outputs. This is based off a chapter in AI for Games by Ian Millington
Dijkstra Pathfinding
this is a pathfinding exercise meant to improve the Dijkstra method by adding a steering behavior, specifically arrival and weighting the different paths so that one is more tactically efficient. For this project I made an AI looking for the most direct path and one looking for the most efficient path. This is based off of projects form Millington's book.
Flocking
This is a game I made to demonstrate flocking behaviors. The small red spheres are lethal and are seeking to the cyan sphere which the player controls to run away from them using WASD and running towards the golden goal. The flocking behavior has an aspect of cohesion which is causing the red spheres to seek each other repulsion to not crash, and steering which makes them go towards the player. This is based off chapters from Millington.
Steering Medley
This is the previous version of the game above and it is another game where the player uses WASD to run to the goal. The 3 steering behaviors used here are Seeking which is the re spheres that hunt the player, Fleeing which has the goal run from the player, and a path follower on the green spheres which creates a patrolling hazard. These are implementations and combinations of pseudocode from Millington.
Ballistics
This is an example of using a mathematical firing solution to determine the strength and elevation of a firing solution in order to get the rigidbody to a predetermined target. I added the ability to add more sequential targets for more jumps and I brought it into VR. The player hits the space bar to launch themselves. This is based on chapters from Millington.
Seek and Flee
This project is a demonstration of the seek and flee steering behaviors which enable an AI actor to track an object, face it and either move to or away from the object. This as accomplished by dynamic forces being added to the current velocity of the actor. This is based on pseudocode by Millington.
