A light-hearted 3D platformer prototype created in Unity, developed for the Games Development module.
The project draws inspiration from classic mascot platformers such as Crash Bandicoot and Banjo Kazooie focusing on exploration, collection, and transformation-based progression.
The player takes the role of a shapeshifter who must regain their powers to lift a curse placed on their wife.
Gameplay revolves around exploring the outback, collecting currency, purchasing potions, and unlocking animal transformations to overcome environmental challenges.
The game was designed to be family-friendly and focused on delivering a compact proof of concept for the genre.
- Third-person exploration with physics-based player movement
- Animal transformations with distinct abilities:
- Kangaroo – high jumping
- Emu – fast sprinting
- Bandicoot – wall climbing / small size
- Interaction minigame using trigger colliders and button prompts
- Dynamic UI and persistent progress tracking via singleton pattern
- Scene transitions for menus and story progression
- Integrated ambient audio and environmental particle systems
- Simple auto-save and load system using JSON
- Engine: Unity 2022.3
- Language: C#
- Tools: Blender, GIMP, Audacity
- Custom Assets: Low-poly animal models, basic rigs and animations, and UI elements created in Blender and GIMP
Environmental assets were supplemented using Broken Vector’s Low Poly Packs for stylistic consistency.
Additional project materials are included in the repository:
- An-Aussie-Adventure_Design-Document.pdf — Initial concept and planning
- An-Aussie-Adventure_Progress-Update.pdf — Implementation reflection and progress report
- Video Link — University Submission— Video demonstration and breakdown
Note: The university YouTube account may close without notice. The videos are retained and could be reuploaded upon request.
This repository represents the final academic version, tidied and verified in the latest Unity 2022.3 LTS release.
Most planned systems were implemented, though several design ideas and stretch goals remain incomplete.
Further development may continue in a separate repository.
Originally completed in 2024 — cleaned and documented in 2025.