Rodot Rodots.

We have twice presented a prototype installation that we call Rodots (aka Lo-fi Collaborative Agent Populations) at the Ontario Science Centre in Toronto, Canada,
in May and July 2007. A broad spectrum of the public interacted with the projection, which began with a phase of "pixel-dust" gathering followed by a brief interlude in which a cluster of pixels followed one tagged participant, and finally transformation into a graphic that enabled the tagged interactant to participate in a herding task. The interactant-representative agent had to collaborate with two virtual agents to herd a target dot into a red circle. The sound for the prototype is based on a bank of oscillators that respond to the position of participants tracked by the overhead camera.


Sample Video Sequences

This was captured directly on thecomputer that was running the program (in Max/MSP, with Java integrated).

How it works
  • We dubbed the first phase of Rodots as pixel-gathering. An interactant gets tagged when they hit a "sweet spot" in the camera's view. At first a cluster of dots follows them, then that turns into one larger cyan dot as the "herding task" sequence begins.
  • Watch the boy in the striped shirt, who appears right after the target goes into the nest the first time. He's very lively, so much so that when an adult tries to coax him to leave later on, he just won't go. He doesn't care whether he has the tag or not. Other people would sometimes try to take the tag away from the person who got it.