OFFER REQUEST MANAGEMENT IN SERVICE ORIENTED HOLONIC MANUFACTURING

摘要

This paper explores the offer request management domain in manufacturing context considering two main directions: costs and time. The goal is to enable manufacturing enterprises to respond in real time to customer requests with offers that are computed in an accurate manner. Traditionally the offer was based on empirical information and often leads to financial losses. The main reason for this is the lack of real time information about the current production line available capacity. Another important aspect when creating the customer offer is the ability to decide on what operations can be externalized to third party suppliers and what operations can be done in-house, in regards to time and costs. This paper argues that the advances in standardization in the manufacturing area, with SOA orientation being the most notable one, help in centralizing the real time production information and allow systems to compute in real time the optimum offer for each customer order. The offer creation problem has two sub-problems: first the producer must decide on which service to externalize and the second one is to compute the overall batch production time and costs. To solve this problem, we propose a genetic algorithm implementation that considers both hard conditions like resource and material availability and soft conditions, like profit, costs and time in which the batch order can be completed. The algorithm is targeted to solve the offer problem in the context of holonic manufacturing systems architectures. The original contribution in this paper consists in the novel architecture proposed for the Offer Request Management module together with the genetic algorithm based implementation. Experimental results are based on a sequence of customer orders for T shaped and Hshapedproduct assemblies and show the differences between the approach based on empirical rules and the genetic algorithm approach proposed in this paper.