GOAP Planner in Unity
I had run into some GOAP implementations in Unity and while all had some good attributes (I even implemented the same FSM model as used by this implementation). I decided to create my own implementation that uses an actual A* pathfinding algorithm (rather than tree traversal) with a regressive planner which is more efficient and intuitive than a progressive planner.
You can view it here on GitHub.
After Pixel is a 2.5D puzzle platformer that I developed in 72 hours for the GMC jam. Inspired by Fez, puzzles in this game are solved by manipulating the 3D perspective in game in order to reach areas.
I learned a lot of lessons in the development of this game. From puzzle development to simple concepts of 3D such as projection.
View more here.
Rhythm of War
I developed a demo for a rhythm game to learn about how rhythm games worked. I wanted the development of the demo to be challenging so I decided to make it an action platformer. This meant that I had to tie the AI of the enemies and the way they the act on their chosen actions to the rhythm in the game.
This was further complicated by the fact that I decided to use a GOAP architecture for the AI. The reason this was chosen was that it allowed for more modular AI agents. I was able to drop in actions and have different AI agents attack the player in different ways.
Initially, I wanted to tie the players actions to the BPM in the game but this proved difficult in an action platformer for the players. Finally, I decided upon only restricting the players shooting mechanic to the BPM and restricting the challenge to timing actions in response to the AI moving to the rhythm.
I learned a lot in this project such as determining when a beat should have been played from the BPM of a track taking into accordance the current time in the song and also how to allow for AI agents to plan using a GOAP architecture.