Goal-Oriented Action Planning (GOAP goap)
Related to Game AI, a newer take on state-machines / behavior trees.
Allows for more emergent behavior.
GOAP supports dynamic prioritization of goals based on world-state.
Also supports nuanced planning building based on cost-analysis.
Often uses A* to build the 'path' (plan).
godot goap impls
other impls
the FEAR paper
goap + finite state machines
resources
Backlinks
Game ai?
To me this means:
But it should probably mean more!
Some of these implementation nodes!
See also: Goal-Oriented Action Planning
See also: beehive
Finite State Machines (FSM).
A classic Game AI pattern.
See also: behavior trees
See also: Goal-Oriented Action Planning
Trying to get a handle on larger game dev topics as I encounter and tackle them.
Game AI can be modeled with behavior trees.
We want a very expressive behavior tree lib, so we can be expressive in our AI design.
Ideally Behavior Trees can be visualized and debugged easily.
Enemy AI design, Level Design. It's all coming together.
Somewhat related: Build a Level editor
See also: Goal-Oriented Action Planning (GOAP)