A real-time fish simulation for a 3D aquarium shown on a spherical display: schooling, obstacle detection and A* pathfinding.
Unreal Engine
C++
Blueprint
AI
Pathfinding
About the project
A real-time simulation of a living aquarium, built in Unreal Engine and rendered on a spherical display. Dozens of fish swim, school and react to their environment, with different behaviours depending on the species.
The client and the project name are under NDA: this page only shows the systems I worked on, captured from development builds.
What I did
Built the whole fish detection and movement system.
Schooling (flocking) behaviour: fish group, align and follow each other while keeping their distance.
Obstacle detection with a 3D spatial grid around the decor, so fish swim around rocks and props instead of through them.
A* pathfinding in 3D through that grid, smoothed into natural swimming paths.
Mapping of fish positions onto the spherical display (latitude / longitude) for the final projection.
Final render on the spherical displaySchooling and obstacle avoidance in the aquariumA* pathfinding between obstacles (editor, greybox)Debug view: fish positions mapped to the sphere
