A supply chain is a network of production and distribution facilities (nodes), typically involving multipleorganizations, that performs the function of transforming input resources (supplies) into finished productsand services delivered to consumers. Supply chain management (SCM) deals with planning and operatingissues of a supply chain. This can be loosely divided into strategic issues such as optimizing the network(e.g., choosing sites for facilities), optimizing products for distribution (e.g., space-efficient packaging anddelayed localization), and collaboration with network partners (e.g., information sharing and risk sharing);and tactical and operational issues such as demand forecasting, order promising, materials sourcing,production and distribution planning and scheduling, inventory control, etc. Relations to suppliers andcustomers (or distributors) are often referred to as vertical, cf. vertical integration. Analogously, relationsto nodes on the same level, competitors and complementary suppliers, are referred to as horizontal, cf.horizontal integration. The famous bullwhip effect is a problem of vertical coordination, where demandsignals are magnified upstream a supply chain in the absence of information sharing. A supply chainmanagement system (SCMS) manages the cooperation of these system components. In the computationalworld, roles of individual entities in a supply chain can be implemented as distinct agents. Fig: Agent Based Model Fig: The Multi Agent ArchitectureDIRECTIVE DECISION DEVICES: Directive decision devices are computer constructs that have beenassigned decision responsibilities that were previously, or would otherwise be, invested in people. Adirective decision device would have to be both autonomous and intelligent, at least to the extentdemanded by the decision situations in which it’s to be employed. Not all decision agents can pass asdirective decision devices. The usual decision agent is designed to perform at the request of, or under themore or less constantly positive control of, a human master (and is likely to be kept on a very short leash).Moreover, a good number of what are promoted as decision agents actually have no decisionresponsibilities, per se. They are, rather, confined to making information-related choices, charged with1ferreting out data items that have some statistical likelihood of being of interest to their principal, and somight then more accurately be described as search agents. Agents operate within organizations wherehumans must be recognized as privileged members. This requires knowledge of organization roles andrespecting the obligations and authority incurred by the roles. Coordination and negotiation must takethese issues into consideration as well, in addition to the computational cost, complexity, and accuracy ofthe algorithms used in optimizationThere appears to be four functional forms in which directive decision devices, actual or prospective, might appear:(1). Executory : There is standardization in pursuit of consistency in the determination of decision choices, designedas a substitute for a human decision-maker, for example: financial ploys like Programmed Trading, whereinvestment decisions are a product of reactive computer algorithms rather than reflective judgment.; (2).Compensatory: Compensatory devices are designed to undertake activities that require capabilities that humans ingeneral or particular may not have in sufficient degree. Compensatory directive decision devices would then betargeted at knowledge-related (substantive or methodological) deficiencies and analytical (mathematical, in themain) shortfalls. Consider, for instance, the computerized income tax preparation programs now so popular in theU.S. They are primarily promoted as a response to the intricacies of the US Tax Code. Yet most of those purchasingthem will have only relatively rudimentary tax issues to resolve. But most of those purchasing them will also haveonly the most rudimentary of arithmetic skills; (3). Interdictive: An interdictive device would be deployed in thehope of preventing prospectively harmful decisions from being enacted. As with compensatory constructs,interdictive devices can be mechanical artifices. These will typically be intended to prevent actions that may beinjurious, or even fatal, to the individual taking them. Anti-lock braking systems, for example are meant to precludethe instinctive human tendency to over-brake in response to skids or impending collisions, an action more likely toexacerbate than avert whatever peril inspired it; and (4). Cooptive: Assume control in cases where humanfunctionaries have failed to initiate an appropriate/timely response to a potentially perilous situation, for example,that might anchor next-generation Traffic Alert and Collision Avoidance Systems. An Agent-Based Model (ABM) is a computational model collection of autonomous decision-makingentities called agents. An agent is an autonomous, goal-oriented software process that operatesasynchronously, communicating and coordinating with other agents as needed. Each agent individuallyassesses its situation and makes decisions on the basis of a set of rules. Agents may execute variousbehaviors appropriate for the system they represent—for example, producing, consuming, or selling.Repetitive competitive interactions between agents are a feature of agent-based modeling, which relies onthe power of computers to explore dynamics out of the reach of pure mathematical methods. At thesimplest level, an agent-based model consists of a system of agents and the relationships between them.Even a simple agent-based model can exhibit complex behavior patterns and provide valuable information2about the dynamics of the real-world system that it emulates. In addition, agents may be capable ofevolving, allowing unanticipated behaviors to emerge. Sophisticated ABM sometimes incorporates neuralnetworks, evolutionary algorithms, or other learning techniques to allow realistic learning and adaptation.It combines elements of game theory, complex systems, emergence, computational sociology, multi agentsystems, and evolutionary programming. The models simulate the simultaneous operations of multipleagents, in
View Full Document
Unlocking...