Robotics in the Evolution of Complexity Science

The evolving science of complex systems consists of many inter-related but incompletely synthesised observations and ideas accumulated over the last century. Some outstanding problems can be identified as major hurdles to be overcome. Trying to understand these ‘bottleneck problems’ in the context of human systems is particularly difficult for two major reasons: it is difficult for humans to study human systems from the inside; and it is often impossible to perform repeatable experiments on complex human systems. It is proposed that communities of robots are complex system, manifesting many of bottleneck problems in complexity science. Furthermore, it is possible to set up systematic and repeatable experiments with robot systems that would be impractical or impossible for human systems. Thus robot systems are simpler than human systems, and they form an intermediate class of complex systems which can, and probably should, be used to investigate some of the most fundamental problems in complex systems science. By providing new insights and answers the bottleneck problems, robotics can make major contributions to the science of complex systems.