Article Preview
TopIntroduction
Manufacturing supply chains are increasingly becoming virtual and agile to meet the need for rapid and cost-effective product development in today’s volatile economy. An agile supply chain can be defined as a network of suppliers that pool their resources to meet short-term objectives and exploit fast-changing market trends (Gunasekaran et al., 2008). The agile supply chain minimizes the excess capacity and capabilities by dynamically adjusting its resources based on the actual requirements of the work orders in hand. The loosely coupled nature of the agile supply chain allows for dynamic addition or removal of the service providers as needed. In addition, agile supply chains are short-lived and typically dissolve once the order is fulfilled. This type of supply chain is particularly suited for Service Oriented Manufacturing (SOM) in which supply and demand are modularly represented as well-defined and self-contained units of service. Deployment of an agile supply chain is challenging for several reasons. First, assessments of the operational and technological capabilities are not readily possible because of the virtual relationship between the actors. Second, a lack of standard models for formal representation of engineering information, particularly capability information, is a hurdle to efficient exchange of information among the participants in the early stages of supply chain formation. Third, human agents cannot efficiently manage the search and evaluation process because of the large size of supply and demand pools. For these three reasons, therefore, it is imperative to support the deployment process with the necessary computational tools and information models that enable automated supply chain configuration and customization with high precision in a short period of time.
One promising solution for addressing the aforementioned challenges is incorporation of agent technology. A supply chain is a natural application domain for an agent-based framework as a supply chain can be considered as a network of autonomous, distributed, and self-contained business units aiming at the procurement, manufacturing, and distribution of goods. Agent-based systems, due to their automation capabilities, can accommodate the computational complexities of the supply chain deployment problem more efficiently. Previously, the Digital Manufacturing Market (DMM) (Ameri & Patil, 2012) was introduced as an agent-based framework that provides the participants, including buyers and sellers of manufacturing services, with advanced computational support for search, evaluation, communication, and negotiation in order to build agile supply chains. Also, a standard ontology was developed and implemented in DMM to facilitate inter-agent communication. The focus of the previous research was primarily on conducting one-to-one matchmaking between supply and demand entities based on their similarities. However, to evaluate the performance of the market and its underlying computational models, it needs to be evaluated in action through simulation. The objective of this work is to simulate the DMM in order to study the behavior of the market under different scenarios. The core functionality implemented in the simulated model of DMM in this work is to build supply chains that fulfill various work orders through dynamic allocation of requested services to the qualified suppliers.
The remainder of this paper is organized as follows. The next section provides background information about the DMM and its associated information model. A review of related works in service allocation is discussed afterwards. The implemented modeling and simulation techniques and their corresponding experimentations are provided next. The paper ends with discussions and concluding remarks.