Doodl is an online drawing robot that feeds on photos. You can give him your own pics, or instruct him to get those for a certain user or tag from Flickr, and he will instantly begin to draw getting his 'inspiration' from them. You can collaborate with him by tweaking several parameters that will change the style and look of the drawing in unlimited variations.
This robot is an interactive tool, a social web application and a generative experiment. It builds on a simple idea: the translation between color and movement. The source images are taken as raw streams of data: grids of numbers that will control the flow of lines in the final drawing.
The robot is actually a swarm: it is made up of a large quantity of autonomous computational agents that drift endlessly over the surface of the screen, leaving a trace as they do it. The source images are like a virtual terrain for them, by means of some simple formulas that turn color values into a force field that influences the speed and direction of their flow. Each agent decides its own behavior, but you can control many global parameters that will have an effect (though often an unpredictable one) on the emergent result of their population dynamics.
In short, you can take Doodl as a creative toy, as a beautiful time-waster or as a serious design and art production tool. The many variables involved open an endless space of possibilities that calls for a lot of playing and experimentation. It is a way to bring new life or extract unexpected beauty from even the dullest image.