Mothia

For the first time I used Miro to document everything before working on them. I feel that it helped a lot with actually putting it into code afterwards. Here we can see the different inputs the director would use to determine how good the player was.

Here we can see the encounter only spawning one enemy because the game doesn't think the player is very good.

Here we see the same encounter as before but this time it has more enemies and even a special stationary sniper at top left. This is because the Director believes that the player is better than previously.

Surprised enemies are vulnerable and can easily be eliminated.

When losing Line of Sight the AI will try to find them by walking to their last known position.

If you are skilled enough bullets can be deflected, this is one of the most effective ways to eliminate the opposition but it will also affect the way the Director looks at you.

Created a custom camera and system for our Twiceler over here.

While the AI alone does not put up a big challenge but when they are huge in numbers they can be quite nasty.

I tried to be extra aware of optimization this time and a couple of things am I extra proud of:
 

 - Level of detail on the AI: The AIs Level of detail changed depending on how close the player was. Everything from update, behaviour tree update to rendering of animation/mesh would be affected by this.

 - A Frame budget for the AI, if it didn't manage to update this frame it would start where it left off the next frame.

Our Behaviour Tree

What I made was it so the tree determines if I should swap my state or not. This is made with a state machine as well. This worked well but I am unsure if instead having a behaviour tree for each states update instead to do both the task at hand and to swap state if needed.