It is increasingly common practice to use remote laboratories as teaching/learning tools, opening a new pedagogical approach and changing the essence, in many cases, of the teaching process (the latest example is MOOC). This expansion of the theoretical applications into a "community of practices" enables the movement of the student's experiences from a "classroom only" environment into remote laboratory networks. The student is introduced into virtual environment, hosted by the Internet, thus him/her is exposed to the uncontrollable, non-linear and unpredictable behaviour (the intrinsic properties of any complex system). These phenomena require the addition to the analysis of the pedagogic effect of the remote laboratory networks, of an analysis of the self-organization process and of its results, just as in the case of complex communication systems. The reason for this approach is that remote laboratory networks are a very small part of the existing complex communication systems, such as the WEB or the Internet. Thus the above complex systems will reproduce, at a different scale, their behaviour. Since the study of "complex systems" is still in its very early stages, by this paper the authors only intended to present the first results of an assessment conducted on an existing group of remote laboratory networks. Only after the principles they are governed by are determined, as revealed by the study, we will be able to discuss their further evolution.