Selecting frameworks for multi-agent systems development for the oil industry

Abstract

Throughout the years, the development of multi-agent systems (MAS) has evolved and many frameworks to support such development were proposed in the literature. Some frameworks did not advance; others have evolved according to the type of usage in either academy or industry. Given the large number of frameworks for developing multi-agent systems, it is important to evaluate these frameworks in order to select which one best suits the development project at hand. There seems to be scanty information and recommendations in the specialized bibliography on development frameworks for MAS to be used in support of supply chain management (SCM) by the petroleum industry in particular. We present in this work a methodology for comparing and choosing frameworks to be used for developing MAS for the oil industry. The methodology includes theoretical and practical aspects. Application of the methodology is carried out in a real case, supply chain management scenario offered by Petrobrás, the Brazilian petroleum company, in automating its planning for draining petroleum products.

Keywords

• evaluation Methodology
• MAS Benchmark
• Multi-agent Systems Framework.