CAP'N - 3D Pathfinding using GOAP

CAP'N is a 3D pathfinding system using a GOAP (Goal Oriented Action Planning) algorithm instead of more traditional A* pathfinding systems.

The system works by simulating steps into the future and predicting what would happen to the Agent using raycasts. At each step, the Agent will try and pick an action to maximize the value provided by the Agent's Goal and avoid colliding with the environment.

This system allows for customizable Goals and Agents. The Goals can be more flexible than a traditional A* pathfinding system, allowing for Goals like terrain following and waypoint orbiting, as well as more basic Goals like moving towards a position.

Using this system, many types of flying Agents can be implemented. The Agents will follow paths that respect their limitations, for example, their radius, acceleration, maximum speed, and climb rate. In the case of Agents such as airplanes, it's impossible for these Agents to travel below their stall speed, as this would result in them crashing. The system can plan paths for these Agents that stay above their stall speed. This behavior allows for very natural and realistic-looking navigation for flying Agents.

The system can be extended to add new types of Goals or Agents to match the needs of your project.

Excellent for Agents such as:

• Rotary-wing aircraft (Helicopters, Quadcopter drones)
• Fixed-wing aircraft (Airplanes)
• Airships
• Flying creatures
• Underwater Agents (Fish, Submarines)
• and more!

If you need support or would like to leave feedback, I can be contacted at [email protected] or on my Discord server.

Links:

Web Demo

Tutorials

Discord