Artificial Life Laboratory, Graz, Austria

 One of the prominent challenges in mobile robotics is to develop
  control methodologies that allow the adaptation to dynamic and
  unforeseen environments. The classic approach of hand-coded
  controllers is very efficient for well-defined tasks and specific
  environments but poor in adapting to changing environmental
  conditions. One alternative approach is the application of
  evolutionary algorithms which need, in turn, easily evolvable
  representations of controllers. In this paper, we investigate one
  promising approach of an artificial hormone system as a control
  paradigm which is believed to be easily optimized by evolutionary
  processes. In a first step of this research, we focus on the simple
  task of collision avoidance. We present a brief mathematical
  analysis of this controller approach and an implementation of the
  controller on a mobile robot to check the feasibility in principle
  of our approach. The task is successfully accomplished and we
  conclude with a discussion of the hormone dynamics in the robot.