In recent years, steering a quality-management system (QMS) has become a key strategic consideration in businesses. Indeed, companies constantly need to optimize their industrial tools to increase their productivity and to permanently improve the effectiveness and efficiency of their systems. To solve such problems, two approaches were developed: the Pareto Analytical-Hierarchy Process (PAHP) and the Multichoice Goal Programming (MCGP) methods. The first integrates the Pareto concept and Analytical-Hierarchy Process (AHP) methods and the second combines the MCGP model with AHP methods. The goal was to determine the best solution while simultaneously verifying multiobjective-optimization functions and satisfying different constraints for a real-world case study. The latter was chosen because it presents a major problem for controlling the quality levels of production lines. A comparative study between the two approaches provides a path for designing a tool for decision support to ensure the effectiveness of a corporate QMS.
In an increasingly competitive international context, companies constantly need to adapt and optimize their industrial tools to increase their productivity by implementing a quality-management system (QMS). Aiming to permanently improve the effectiveness and efficiency performance of organizations, a QMS can be measured against International Standards Organization ISO 9000, 9001 and 9004 [1], [2] and [3]. Piloting such a system refers to the general aspects of quality management and its implementation [4].
In this context, several researchers have incorporated the optimization concept in various industrial fields such as the scheduling of complex system maintenance [5], container management [6], dynamic scheduling in the food-processing industry [7], the management of FIR (finite impulse response) filters [8], the assignment of antenna frequencies in the telecommunications field [9]. These practical applications have shown that optimization can yield benefits to companies in terms of its impacts on their resources and economics, as well as productivity gains [10].
Generally, application of this concept to real-world cases with often conflicting criteria is based on a multiobjective function [11]. Several solution methods have been proposed for this type of multiobjective-optimization problem. There are those that convert the multiobjective problem into a single-objective problem [12], [13], [14] and [15]; others have mainly been based on simple genetic algorithms, and the only differences are in how the algorithm selection is made [16], [17], [18] and [19]. A third class of methods is founded on the dominant Pareto concept, which aims to enable a best study that satisfies the objectives [20], [21], [22] and [23].
The difficulty of using these different techniques is ignorance of the relative importance of the criteria. Moreover, their complexity stems from the multitude of quantitative and qualitative criteria influencing the decision choice. An analytical approach often suggested to solve a complex problem is the Analytical-Hierarchy Process (AHP) method [24]. Several researchers have used AHP as a standalone methodology to make location decisions [25], [26] and [27]. Many other studies have used AHP in combination with mathematical programming in disciplines other than location [28], [29], [30] and [31]. The AHP method provides an ideal classification for decision support, but it does not take into account constraints that exist in the decision environment. In this context, and to solve a multiobjective-optimization problem, two approaches were proposed based on the combination of AHP methods with the Pareto concept (called PAHP) and with Multichoice Goal Programming (MCGP).
The paper is organized as follows: Section 2 presents the problem definition. Both approaches are summarized and detailed in Section 3. In Section 4, we present a real-world case study to which we apply the two approaches, followed by a comparative study between them. Finally, we provide conclusions and relevant perspectives.
The piloting of a QMS represents the general aspects of managing an organization that determines the quality policy and its implementation. For this, each company needs to constantly optimize its industrial tools to increase productivity. In this context, we proposed two approaches, PAHP and MCGP, for solving this kind of MOP. Both approaches were used to construct tools for decision support to ensure the effectiveness and efficiency of a QMS in the case study. Moreover, these two approaches were considered appropriate with respect to the requirements of the ISO 9000, 9001 and 9004 standards; these ensure the objectives of a QMS such as customer satisfaction and system efficiency.
The methods for decision support, such as classification techniques, for example, Analytic Network Process (ANP) [42], the metaheuristic techniques [43], mathematical models such as Multisegment Goal Programming (MSGP) [44] and the integration of fuzzy logic [45] in assisting human judgment, present significant opportunities to help the decision maker to make the best choice.
