Modelling Production Management Problems by Knowledge-Based Approximations

Abstract

Production Management (PM) in a discrete multi-product manufacturing enviroment consists in deciding, for each order of final products from customer,

• (A) when the manufacturing system must initiate to process items

• (B) by what machining center (or sequence of machining centers) the items must be processed

• (C) in what rates the working time of each machining center must beshared and assigned to items of different product families

Formally the PM problem can be started in terms of a dynamic scheduling problem, for given due date constraints of orders from customers and for given release constraints of material from suppliers.

A formal solution of this problem, suitable for a practical use in a multi-products multi-center industrial department, has not yet been found.

Approximation is mandatory. But it is not an easy task, since an approximated solution of a combinatorial optimization problem must be found, together with some criteria to evaluate the approximation accuracy.

In author’s opinion this problem can be approached by using the wide experience of industrial managers, thus adopting knowledge-based criteria from production management practice.

The contents of this contribution will be focused on such an approach, first introducing a K-B approximation methodology, then analyzing its applicability in solving PM problems.