Software Agents Oriented Approach To Warehouse Management Systems Design

摘要

The last two decades have witnessed great changes in the way the processes related to goods storage, distribution and delivery are managed and controlled. New concepts like Just-In-Time (JIT), Lean Manufacturing (LM), and Supply Chain Management (SCM) have imposed many requirements on warehouse management systems (WMS) related to performance improvements, including reduced cycle time, faster order delivery, more accurate inventory, customer predelivery information, etc. The Internet boom had a big impact on the goods storage and distribution as well. It sharply increased the volume of received orders and customers' expectations. One could say that all warehouses and autonomous Third Party Logistics (3PL) business systems started facing the old challenges with a new magnitude: how to control the movement and storage of goods and record movement transactions with the greatest possible accuracy and lowest possible labor costs while maintaining optimal customer service. At the same time the architecture of WMS systems has not been changed neither to adopt Just-In-Time (JIT) and Lean Manufacturing (LM) concepts nor to efficiently respond to the effects of the Internet boom. Current designs of WMS systems, whether they were results of the traditional or object oriented analysis, are not an adequate answer for the existing challenges. A new design of WMS systems has to offer such solutions capable of monitoring events in the warehouse 24 hours daily and execute appropriate actions. In other words, a new WMS system design has to have a goal oriented processes as opposed to traditional task oriented processes. Framework for such development of WMS systems is an agent-oriented analysis and design. In this paper application of agent-based modeling to the management and optimization of the warehouse operations is presented. It is an attempt to develop the hardware and software architecture that shall integrate and orchestrate all these subsystems and interconnect them, through distributed software agents, with a centrally operated database and a middleware system implementing all the business intelligence behind the control logic and decision-making scenarios serving business objectives.