Projects

These are some examples of projects I have worked on during my studies at AIE. I was also grouped with others from AIE. 

    Overboard
Rapid Delivery 
Custom C++ Engine
Machine Learning AI

Overboard

Systems Developer

Overboard is a game that I produced as a student in my second year of studies at AIE, in a group of 8. We spent 6 months working on the project in total, planning and documenting the process alone the way. 

I focused on the creation of systems for the game, that not only were used in gameplay, but were also used by artists and designers in the creation of the game. My main goal when creating the systems was to make them modular in order to increase the speed and efficiency of development. I also focused on making the system easy to use for artists and designers, they were able to easily call functions, input variables and broadcast global events. The modular system only required a minimal understanding of coding which I was able to teach my team members and the were able to easily tweak animations, cut scenes, character controllers, boss battle settings, etc.

Overboard
    Rapid Delivery 
Custom C++ Engine
Machine Learning AI

Rapid Delivery

Systems and Gameplay developer

Rapid delivery is a game that I created also as a student in a team of 6. The project was given a brief, from a tea company requesting to create a game to promote their new tea collection. 

For this project, I worked on most of the visual and playable aspects of the game including the environment spawner, obstacle manager, player controls, shooting mechanics, etc. We worked with existing prefabs in the Unity scene in order to make the environment spawning process as efficient as possible. 

Overboard
 Rapid Delivery 
 
    Custom C++ Engine
Machine Learning AI

Custom Bootstrap Engine in C++

During my studies we also developed a custom bootstrap based game engine in C++. We wrote most of the engine code aside from window deployment, including physics, shaders and graphics processing with both 2D and 3D support

Overboard
Rapid Delivery 
Custom C++ Engine

    Machine Learning AI

Custom Machine Learning AI in Unity

During my studies we were also tasked with creating a complex game system of choice from scratch. I decided to research how machine learning AI's work and create my own from scratch. (I did not use Unity's MLAgent program for this project)

 

The program first allows the user to create a path with an agent. Once a path is found it can be saved to file and then used by a bot which reads the information and plays through the path automatically. This bot can be played against by players.

The program does feature a series of limitations and potential optimisation points of which I would like to go through at some point. These issues are in place because there is very limited information on how to create a Machine learning AI. Though there is lots of information about how AI's think, the actual calculations are often referred to as "hidden layers". Though I was able to create my own very basic machine learning AI which does it's job and of which I learnt a lot from.