Controlling cooperative and conflicting continuous actions with a Gene Regulatory Network

Artificial Gene Regulatory Networks (GRN) usually simulate cell behavior in developmental models. However, since 2003, GRN based controllers have been applied to robots to solve problems with few sensors and actuators. In this paper, we present our first steps toward an effective GRN-based controller for intelligent agents in video games. We will also introduce an experiment, the Radbot, where a robot has to handle and manage simultaneously four conflicting and cooperative continuous actions. Finally, we will show how a GRN-based controller can be evolved to solve the Radbot experiment.