Underwater World
Angelina Zhumadilova, Ander Carrasco Del Rio, Sanna Mitt
Windows built (google drive): https://drive.google.com/file/d/1Il1ymgaRhxOwJVh_g0WPBu4nOyW5F6DD/view?usp=sharing
Repository link: https://github.com/angelinazhuma/ProceduralUnderwater_Project
Description
This project is a real-time underwater environment created in Unreal Engine 5. The goal was to design a visually immersive underwater scene using modern engine tools, procedural generation, and real-time visual effects, while focusing on atmosphere rather than gameplay.
The scene represents a stylized underwater world with natural terrain, vegetation, rocks, corals, and animated fish, enhanced by light caustics and volumetric light rays.
What Was Done
- Procedural placement of underwater plants, rocks, and corals using the PCG (Procedural Content Generation) system.
- Creation of an underwater terrain and environment layout.
- Implementation of animated fish using Niagara particle systems.
- Setup of water caustics and light rays to simulate underwater lighting behavior.
- Adjustment of post-processing effects to achieve underwater color grading and depth feeling.
- Implementation of autonomous crab movement using AI navigation enabling natural wandering behavior.
- First-person camera mover (you can explore around with WASD, space and mouse).
How It Was Done
Procedural Environment
PCG graphs were used to automatically distribute plants, stones, and corals across the environment. This allowed fast iteration, natural variation, and scalability without manual placement of assets.
Fish Movement
Fish behavior was implemented using Niagara Hanging Particle Systems to create believable, ambient motion throughout the scene. Three types of fish schools were designed, each using a combination of Vortex Force and Point Force. The blue fish and larger fish follow smooth, circular paths driven by stronger vortex forces, reinforcing a calm, continuous flow in the environment. In contrast, the clownfish school features a more erratic and localized movement pattern, achieved by reducing both vortex and point forces.
Lighting and Atmosphere
To enhance realism and immersion:
- Light rays were added to simulate sunlight penetrating water
- Caustics were applied to surfaces to mimic light refraction
- Post-process volumes were used for color shift, contrast reduction, and underwater haze
Crabs walking
Crab movement was implemented using Unreal Engine’s AI navigation system. The underwater terrain was configured as a navigable surface, allowing the crabs to understand which areas are walkable on the seabed and which are not.
Each crab periodically selects a random reachable point within a defined radius around them. The crab then moves toward this location using AI navigation, creating a continuous loop of autonomous movement. Once the destination is reached, a new target point is selected, resulting in natural, wandering behavior across the terrain.
Distributon of task (as of November 11):
Angelina Zhumadilova - project setup and environment creation, building the Unreal Engine Project and managed the courses page
Ander Carrasco Del Rio - made windows built, participated in developing the enviromental idea, helped manage the project’s course page
Sanna Mitt - captured illustrative screenshot, participated in developing the environmental idea, helped prepare materials for the milestone submission