AN AGENT-BASED COLLABORATIVE MODEL FOR SUPPLY CHAIN MANAGEMENT SIMULATION C. M. Vieira A. P. Barbosa-Póvoa C. Martinho Centre for Management Studies Centre for Management Studies Dep. of Computer Science Technical University of Lisbon Technical University of Lisbon Technical University of Lisbon 1049-001, Lisbon, Portugal 1049-001, Lisbon, Portugal 1049-001, Lisbon, Portugal ESTG E-mail: apovoa@ist.utl.pt Email:carlos.martinho@ist.utl.pt Polytechnic Institute of Leiria 2411-901 Leiria, Portugal E-mail: carlos.vieira@ipleiria.pt KEYWORDS Supply Chain Management, Multi-Agent System, Simulation. ABSTRACT In traditional supply chain (SC), planning problems are usually considered individually at each SC entity. However, such decisions often influence the other members in the chain and thus an integrated approach should be considered. By modelling system-wide SC networks, different SC problems, like production planning, coordination, order distribution, among others, can be integrated and solved simultaneously so that the solution is beneficial to all entities in a long- term base. In an attempt to make progress in this area, researchers use various methods for modelling the dynamics of SCs. In the literature review, due to their distinctive characteristics, multi-agent-based systems have emerged as one of the most adequate modelling tools for tackling various aspects of SC problems. In this work, a multi-agent supply chain system (MASCS) model that integrates different SC processes is presented. The proposed model allows modelling different SCs with multi-products and different operational policies considering information asymmetry and distributed/decentralized mode of control. In this article the details of the MASCS model development and implementation are presented. Furthermore, the applicability of the proposed MASCS is briefly demonstrated through the solution of a SC example. The obtained results are discussed and research extensions are outlined. INTRODUCTION A supply chain (SC) is a network of trading partners linked through upstream and downstream connections where the main aim is to produce and deliver products/services to the ultimate consumers so as to provide global SC profit. Traditionally, managers have been focusing on the management of their internal operations to improve profitability and thus an internal concern has been the main objective. However, supply chain management (SCM) calls for the integration of the SC operational activities in order to organize/manage the flows between entities as if they form a single organization. Moreover, with the businesses globalization, inter-organizational coordination is becoming strategically important for companies to augment responsiveness while maintaining SC efficiency. Numerous studies have demonstrated that substantial benefits can be obtained from an integrated SCM. Such integration provides tremendous challenges to managers (Arshinder 2008). Although a completely integrated solution may exist with an optimal system performance, such solution is not always in the best interest of every individual member. As a result, each SC member attempts to optimize a part of the system without giving full consideration to the impact of their myopic decisions on the total system performance. Optimizing the portions of the system yields sub-optimal performance, resulting in an inefficient allocation of scarce resources, higher system costs, compromised customer service, and a weakened strategic position. A key issue in SCM is then how to coordinate the independent players to work together as a whole so as to pursue the common goal of chain profitability. In an attempt to make progress in this area, researchers use various methods for modelling the dynamics of SCs. The multi-agent system (MAS) approach has appeared as one of the most adequate modelling tool for tackling various aspects of SC coordination problems (Santa-Eulalia et al. 2011). Indeed, considering the fact that most of the SCs involve enterprises with independent ownerships (requiring the ability to model information asymmetry and distributed/decentralized mode of controls), applicability of traditional modelling approaches is limited and unrealistic (Govindu and Chinnam 2010). Additionally, a holistic model is required to explore different coordination mechanisms and their value in SC where different set of combinations of coordination mechanisms can be tried with the help of simulation. Moreover, most of the models describing coordination mechanisms are dealt in two-level SC, which needs to be extended to multi-level SCs (Arshinder 2008). Proceedings 26th European Conference on Modelling and Simulation ©ECMS Klaus G. Troitzsch, Michael Möhring, Ulf Lotzmann (Editors) ISBN: 978-0-9564944-4-3 / ISBN: 978-0-9564944-5-0 (CD)