Revisiting Reynolds - Autonomous Agents for Spatial Audiovisual Composition and Performances

Autonomous agents represent a special class of systems in the context of Artificial Intelligence (AI) and Artificial Life. As embodied AI, they are systems that have a certain understanding of their environment and adapt their behavior accordingly. They are used in a wide range of fields, from robotics to video games, but also in artistic practice. Craig W. Reynolds' work on steering behavior and flocking simulation provided an important framework for simulating the motion of autonomous agents. Originally intended for use in computer games, various adaptations and inspired applications can be found in a wide variety of domains, including spatial composition. The motion of autonomous agents in 3D space can be used to control spatial sound sources and other virtual objects to realize life-like behavior in improvisations or conducted performances. In this paper, we describe the implementation of steering behaviors proposed by Reynolds as autonomous agents in the modular 3D engine IVES for the Max development environment. Furthermore, we demonstrate the potential to realize spatial audiovisual compositions and performances with the improvisational behavior of autonomous agents in combination with a 3D engine specialized for art production.