Slime Herder

What is it?

Slime Herder is an experimental spawning system visualised through gameplay. it is a Top Down simulation experience in which the player must herd all of their slime into the nest before they can be eaten by predators.

What can it do?

The spawning system allows up to 1000 AI entities to be spawned on screen at any one time while maintaining up to 30fps at maximum spawn amounts. The spawn system considers that AI might need to be complex in some cases and allows for the creation of interesting AI systems, spawning them without sacrificing performance.

How does it play?

The gameplay loop is simple in which the player has to herd their slime creatures into the nest in the middle of the arena before the predators can eat their herd. The player can attack the predators to help relieve some pressure but their main objective is to get the slimes home!

The game features three types of AI creatures, each with unique behaviours.

• Pack Leader

The pack leader, leads the pack which in this case is a huge herd of slime! The pack leader can be manoeuvred around the map by the player as the pack leader will always try to run away from them. Using this to the players advantage, the player can herd the leader into the slime nest where all of the pack creatures will begin to follow to.

• Pack Creature

Pack creatures, led by the leader are a huge swarm of slime. They will always try to follow their leader while simultaneously trying to avoid the player. Like the pack leader, pack creatures will attempt to run away from the player, allowing the player to usher the pack creatures into the nest. 

• Predator

Predators will spawn at intervals set by the player at the start of the game and immediately begin to target the players herd. Predators will indiscriminately attack the players pack creatures and eat them!! The player can fight back to buy some time but their main goal is to safely bring as many slime home as possible.

Controls 

• WASD to control the character
• E to attack predators
• P to pause the game

_______________________________________________________________________________________

This system is still under development. The current AI systems are temporary and can be optimised further by transferring behaviours from blueprint to state trees or behaviour trees. C++ can be utilised to further optimise the AI.