Agent-based modeling framework for complex adaptive organizations

Abstract

Even though humans achievements are due to their particular way to reason about the world, with approximations, simplifications and derivations, complexity always amazed us, with its deceivingly simple base rules and unforeseen consequences. Many are the systems of individuals found in natures in which the result of of their activities is higher than the sum of its parts, as is the example of ants behavior when searching for food: their search method is very efficient, even though each individual only obeys a simple restricted set of rules. Broadcasting companies find tackling the complexity that arises from their environments hard to analyze and comprehend. As such, much of the potential for improvement is out of reach of existing tools. In this context, this thesis aims to provide a tool that helps to reason over the behaviors and interactions that take place in a multimedia production environment. By answering the questions "what is happening", "what will happen", and "what would happen if", the knowledge gathered simplifies the system of interactions in such a way that insight can be harnessed, and therefore, action can be taken. To accomplish its goals, this work subdivides into development of a generic framework and an instantiation of it. The generic framework is constituted by multiple extensible modules, and aims to be instantiable in environments where analysis of behaviors in complex systems is relevant. Thus, not only the example of media production environment is used, but the cyber-physical systems also help to showcase the general scope of the framework. The instantiation of it binds the tool to the specific environment, and is to be deployed alongside a tool already used by many major broadcasting companies.