KARL SIMS

http://www.?

Intelligent virtual environments

A number of new research efforts are emerging that will allow the development of intelligent virtual environments and greatly enhance the experience of telepresence. One major area of research aims to make virtual environments more dynamic and to enable virtual objects to interact with each other. This is the context of Karl Sims’ work to create computer-generated spaces in which virtual elements evolve on their own in accordance with pre-established rules: one example is a landscape in which plants appear and grow spontaneously. A similar direction is being explored by Mike McKenna and colleagues at the Massachusetts Institute of Technology, who are doing research into images of behavioural patterns. Lastly, it is worth mentioning the work of Susan Amkraut and Michael Girard on the animation and simulation of the movements of articulated animals. For example, in Eurythmy, a set of complex algorithms enable birds in flight to cluster together and move individually in reaction to the movements of their fellow creatures. In Menagerie, as soon as a user moves into the virtual space, virtual animals respond to him in real time.

	1995, Evolved Creatures
	1993, Genetic Images
	1991, Primordial Dance
	1991, Interactive Evolution of Dynamical System
	1988, Particle Dreams

1995, Evolved Creatures

These Creatures, modeled according to behavioural rules, call on research projects on learnig patterns and artificial life. Here, the cretures are learning to walk.

1993, Genetic Images

The computer presents the visitor with sixteen pictures which are very simple at first. The computer then removes all those that are not selected by the viewer and generates new pictures which are offspring with mutations and combinations from the ones selected by the viewer. Massively parallel computations allows a new dimension of computing speed, adding more processors that can work all together. And this in turn allows new and different kind of algorithms and interactive process to be tried out that could not previously be performed.

1991, Primordial Dance

The computer introduces random modifications to the equations which generate the forms, but it is the operator aesthetic criteria that control theevolution of the image from one generation to the next.

1991, Interactive Evolution of Dynamical Systems

The computer displays nine populations of images, some of which, after selection, can undergo complex trasformations, thanks to equations employes. Based on the evolution of networks of cellular automations, the emergence of new behaviour patterns in dynamical systems of ever greater complexity is evaluated here by interactive methods.

1988, Particle Dreams

The appearance and behaviour of different particle systems are controlled by giving each particle its own virtual processor within the supercomputer. In this way, systems like these can account for complex physical phenomena, such as fluxes, in realistic or system-enhanced ways. In the figure below the image of a three dimensional face made up of particles decomposes and is recomposed according to predetermined physical laws.