Project created for the game development course taken in SUPSI. The main idea for the game is to have a spherical world where the player can move, with gravity towards its center. While on the planet, the player (that is a sphere like the planet) will be attacked by cubes: stranges shapes that are attacking your planet. The player can defend itself by shooting a raffic of laser shots that will destroy the cubes. The goal is to survive as mush as possible.

Simply go to the releases page, download for your platform and run.

Clone the repository and open it in Unity. The project was created using Unity 2020.1f

This game contains a custom multi threaded implementation with path smoothing of the A* algorithm. A* algorithm is used to achieve a compromise between speed and precision. It usually find a good solution that is indeed not always the optimal solution.

The reason why I devoleped this implementation was to achieve pathfinding on a sphere, which is the terrain of the game. Unity built-in pathfinding was of no help because it only can handle flat or partially flat terrains.

The process basically consists of creating a virtual copy of the sphere mesh, with higher level of detail (usign an Icosphere), and to use each vertex in that mesh as a node in the graph to be feeded to A*. After the terrain is initialized, if an entity capable of pathfinding (i.e. an AIAgent) enters the surface of the sphere, it starts calculating the path to the targe on a separate thread using C# Tasks. When a valid path to the target node is available, the entity starts moving towards his path and in the same time it starts simplifying the path with a naif algorithm.

Nodes that are not traversable are found by casting a ray along the normal of the terrain to listen for collisions with obstacles. In the preview, you can see the path calculated by entities in cyan, and the nodes that are not traversable as red balls.

Further features can be added to the algorithm to improve performace, for example, I planned to add shader computation for the graph to significantly increase the performance of the algorithm.