Collective Behavior Model interface

Here’s an update on the model. As a convenience for the purposes of debugging and experimentation, I’ve added a “dashboard” readout that can be used to display any variable. I’ve also added manual controls to adjust the behavior parameters while the simulation is running. It’s fun! You can cause the agents to clump together, disperse, or exhibit other apparent patterns. Lastly, I’ve added a rudimentary pause mode (bullet time, as I announced to all within earshot upon implementation).

Pausing is handy, but I expect it will need to be rewritten in a way that allows more continuous control of the simulation speed – slow motion or step-by-step modes would also be nice.

Still lots to work on and review in the model itself. In particular, I want to examine the turning rate code; I see a lot of jittery motion that suggests something may be off.

Want to see and hear more of this project? I can continue to post updates. Some will be more focused on the concepts of the model and some will be more focused on the implementation – for example, I don’t really know much about Java, so I may post code examples of whatever strikes me as clever or confusing.

Posted on Thursday, May 13th, 2010. Tags: , , , .

2 Responses to “Collective Behavior Model interface”

Posted by Ian Smith on Friday, May 14th, 2010 at 9:28 AM.

I would like to see your flocking project expanded to include LEGO behavior patterns. For instance, larger LEGO objects will react differently than smaller LEGO objects, and vice versa.

Posted by Jim on Friday, May 14th, 2010 at 12:06 PM.

Perhaps I could make self-assembling models by making the tops of bricks attracted to the bottom of others!