MISSION BLUE

A collaborative project about ocean conservation and proper recycling practices, in line with the United Nations's Sustainable Development Goals.

Our work won 1st place in the British Computer Society AGD SG Artefact 24 competition "AI for Good" and was the best game article submission of the iLRNFuser 2024 Game Jam. The paper can be read here.

The demo is a 3D game made in Unity. The user can complete missions or explore the water species with the help of a conversational AI, "ConvAI."

My role in the project included researching and designing the environment and missions, as well as implementing a recycling mini-game.

During the demo's development, we researched marine species native to the Pacific Ocean to highlight and bring awareness to the affected ocean life living in the polluted waters close to the Great Pacific Garbage Patch. Some of the information was also used to create ConvAI profiles of the animals and the player can learn more about the species by discovering in the water. Discovering the species gives a reward.

The mini-game is a recycling mission. The player collects a certain amount of garbage debris: plastic, glass, or metal. Upon collecting all the required debris, the player must properly recycle them to collect points, which can be used to upgrade their personal AI assistant. The AI assistant can then be deployed underwater to assist with the ocean cleanup.

DUAL

WORLDS

Unreal Engine with Blueprints

For my final year project, I wanted to create a small immersive world where the player experiences a story through dialogues with other characters and interactive surroundings. There are 4 areas: the starting zone, which is a cave, a small village, a forest and a castle.

Each zone introduces a different game mechanic - collecting artefacts, solving a puzzle, a portal to the past and depending on the player's choice - 3 different endings.

Quest System and Dialogue System

The story follows a non-linear approach, which gives the player the power to choose the game's ending. Each NPC has a dialogue tree, that includes several conditions and events that control the flow of the dialogues and depending on the player's actions and choices, it will display the appropriate text.

There is a reputation system, which determines the story's outcome.

If a player finishes a quest, the quest system will reward them with reputation points. The player's answers in a dialogue can either give or take away reputation points and if a player answers in a rude manner - the NPC will give more obscure anwers.

Character Design

There are 2 types of NPCs - quest givers and quest helpers. The quest givers maintain the flow of the main story, give missions and have background stories, while the quest helpers can provide clues to the player if they are currently on a quest. Both types have a parent blueprint and can be customised depending on the role of the npc. Each npc also displays a different level of animation complexity thus expressing different ranges of emotions as a result of a player's choice.

Game Artificial Intelligence Project

This project features a procedurally generated terrain that can be adjusted in run time, enemies with 2 different finite state machines with A* pathfinding and an asset spawner, that adapts to the dynamic terrain.

Procedural Terrain Generation

Enemy Finite State Machine and A* Pathfinding

Battle Royale Game

A multiplayer game, where up to four players can join a lobby, chat both in the lobby and in the game, and engage in fast-paced combat, racing against the clock to determine the ultimate winner.

Maki Mates - Restaurant Simulator

Interactive Media Product showcasing interaction and feedback via dialogue system, responsive characters FSM system and animations, branching narrative, Unity UI system