A New Kind of Art: The Robotic Action Painter

In the context of robotics RAP [Robotic Action Painter] demonstrates that machine creativity can be achieved by means of randomness, stigmergy and chromotaxis. RAP’s action painting is based on two distinct behavior modes, the first being essential random and the second reactive or of positive feedback. The initial random color spots functions as a seed for the reactive mode emergent composition. In this sense the paintings are always distinct and unique. Mode activation is based on local information gather by the robot itself. RAP uses a set of nine RGB sensors turned to the painting plan in order to determine the presence, shape and intensity of color. If color – with considerable expression (threshold) – is found, RAP changes from Random to Reactive Mode and proceeds to final composition arrangements. In this mode the robot is both attracted by color (chromotaxis) and responds to it (stigmergy). The nine RGB sensors disposed in a grid of 3x3 make also possible for the robot to decide when the painting is finished. The fact that RAP can read not just single color spots but also local patterns, permits to generate a kind of “sense of rightness” and let the robot determine by itself the end of the process. RAP is also able to sign its works. Although thresholds are used, essentially for variation of color sensibility, no use is made of any kind of fitness, other predetermined behavior loops (except for the signature), composition or aesthetical targets. RAP was conceived for the largest robot autonomy and consequently the less human intervention as possible. RAP’s paintings are fairly original to be considered the product of machine creativity. Sensors allow for an effective incorporation of new and non predetermined data which is the base of the creative process. Hence, these artworks are founded on the machine own interpretation of the world and not on its human description.