Análise da simbiose industrial por meio da simulação baseada em agentes: aplicação no setor agroindustrial

Abstract

The environment needs beneficial actions to eliminate or reduce the environmental impacts caused by human beings, considering their population increase on a finite planet. Industrial Symbiosis (SI) is focused precisely on the non-generation or reduction of waste discarded in nature, where it seeks efficiency in the use of resources and its constant reuse among different companies. The objective of this thesis was to analyze the development of SI through Agent Based Simulation (SBA) in an agro-industrial system in the city of Dourados-MS considering both the exchange of solid waste and the relationships of agents. In order to increase the organization and structuring of a system with favorable aspects for the IS project, design science is used to assist in this construction, with the understanding of the elements, functions and properties. In addition to the science of the project, the new system needs verification and support in decision making, as every change is highly impactful, which is achieved through modeling and simulation of the variables and relationships existing in the system, in this case considering the actions taken as a basis by the agents. The research object was the city of Dourados-MS for access to data and number of companies. A systematic review on the IS was carried out to understand its characteristics and activities, in addition to an exploratory review on project science. In addition to these, systematic reviews were carried out on the relationship between SI and SBA and SI the object of study in the agroindustrial sector. From the structure to develop the SI and from the modeling and simulation research procedure, the case was understood with visits and electronic contacts with the companies and the construction of its computational model with the equation of the data. With this, the computational model was translated into the Netlogo software, which is most mentioned in IS research and open for use, where simulations were performed to analyze the data and understand the agents' behavior. From the data collected it was possible to identify information about two main contexts, the physical, with the generation of waste and the specificities and properties of the system, and the social, with relationships and influence of social actors in decision making. Parameters were defined to compare the simulation scenarios and identify the best paths for the IS to succeed in the system, such as eco-connectivity and the amount of reused waste. The simulation made it possible to identify the behavior of each variable, as well as the reuse of waste and the number of agents to be created for this, the integration, the relationships between the agents and the positive impact of the presence of the coordinator. The role of government, society, the market, and financial networks also contribute to the evolution of agents, as well as the institutional capacity of each member of the system. A script for structuring the computational model was presented and the applicability of the SBA to develop SI in the presented system was proven, showing the specificities, such as the seasonality and variability of agriculture and livestock.