Swarm system with Grasshopper+Ghowl+Kangaroo
Swarm system with Grasshopper+Ghowl+Kangaroo by Matthijs la Roi
There are many examples of agent based swarm systems developed with the main rules of Craig Reynolds boids. Processing proved to be a powerful tool to produce some interesting swarm systems.
My idea was to develop a swarm system based on Craig Reynolds boids without having to code anything but only use default grasshopper components and plugins for grasshopper.
The problem of grasshopper that by default it isn’t structured to perform the loop based calculation to update every agent of a simulation. From previous experiments with gHowl I noticed that gHowl was capable to send internal messages to your internal IP address. Cause the message have delay set by a grasshopper with frame rate with a minimum of 20ms, looping iterations are possible.
Error what you get if you try to loop some data in grasshopper.
After I managed to model the basic boid system of Separation Alignment Cohesion with grasshopper vector components and kangaroo to recalculate velocities, I started making more simple rules.
In the simulation shown in the animation you see that some agents are leaders. If the other swarm agents are in the proximity of a leader they will belong to this group. Then grasshopper calculated how many group members each group has and will give them color.
This is still a premature study but I will develop the system to form more complex architectural formations of environmental and functional aware agents.
Example of a simulation with the group swarm script.
The system after multiple iterations complex patterns start to emerge.






0 Comments.