Meta-Level Scalability Issues in Analysing the Scalability of Self-Organisation

Addressing the challenge of scalability in self-organisation is ostensibly a “numbers game”. It is concerned with stress-testing a means of self-organisation (e.g. by emer-gence, or planned emergence, etc.) as the number of agents or organisations grows, time-scales (e.g. for decision-making) shrink, and/or the number of decision-making areas (to address inter-dependent games) increases. However, studying the scalability of self-organisation using multi-agent simulation presents three meta-level scalability issues, including platform scalability (scale impacts time/space resources linearly), experimental scalability (fixing some parameters so others can be scaled), and experimenter scalability (the user interface should support scalable programming and configuration). In this paper, we introduce a complex scenario which has multiple dimensions of scalability, and then describe a generic multi-agent platform, consisting of a base layer and a specific instantiation for analysing scalability in this scenario. We briefly review some initial experiments conducted with the platform. Moreover, we offer the code open source and available to others to use the base layer to instantiate their own scenarios. or explore further the one described here.