A fuzzy AHP-based simulation approach to concept evaluation in a NPD environment

Abstract

The evaluation process of conceptual design alternatives in a new product development environment is a critical point for companies who operate in fast-growing markets. Various methods exist that are able to successfully carry out this difficult and time-consuming process. One of these methods, the Analytic Hierarchy Process (AHP) has been widely used to solve multiple-criteria decision-making problems (i.e., concept evaluation, equipment selection) in both academic research and in industrial practice. However, due to vagueness and uncertainty in the decision-maker's judgment, a crisp, pair-wise comparison with a conventional AHP may be unable to accurately capture the decision-maker's judgment. Therefore, fuzzy logic is introduced into the pair-wise comparison in the AHP to compensate for this deficiency in the conventional AHP. This is referred to as fuzzy AHP. In this paper, a fuzzy AHP method is used to reduce a set of conceptual design alternatives by eliminating those whose scores (or weights) are smaller than a predetermined constant value obtained under certain circumstances. Then, simulation analysis is integrated with the fuzzy AHP method, and the hybrid method is used to help the decision-makers (product engineers or managers) evaluate the remaining alternatives from the fuzzy AHP method. A real-life manufacturing system is used as the testbed for the proposed techniques. Finally, the results of both techniques, fuzzy AHP and simulation, are used for Preference Ratio analysis to reach to the final alternative.

Acknowledgements

The author thanks Frank Kowal, Senior Engineer, and a cross-functional project team especially set up for this study at Mold-Master Ltd., Georgetown, Ontario, Canada, for their contributions to this paper.

Notes

¹Fundamental scale used in pair-wise comparison (Saaty, 1989).

¹Set-up cost (CDN$) = average 0.05 h per unit × CDN$40/h × the number of operation-centers (1) in the route matrix for each part of any concept [58 = 0.05 × 40 × 29]. Set-up costs are assumed the same for any part of each concept at any operation-center.

²Process cost (CDN$) = (total process time cycles in all the operation-centers for one unit of each concept) × (process cost for each operation-center—CDN$ 45/h) [329.85 = 45 × 7.33 h for Concept A; 387 = 45 × = × 8.60 h for Concept B and 379.8 = 45 × 8.44 h for Concept C) Processing costs include overhead charges.

³Material cost (CDN$) = (weight-kg for each concept) × (cost of material per kg). All material cost per kg (mostly stainless steel) is approx. CDN $22 per kg [Concept A, 401.28 = 22 × 18.24 kg; Concept B, 424.16 = 22 × 19.28 kg; Concept C, 403.48 = 22 × 18.34 kg]

*Industrial cost per unit is assumed to be the sum of the set-up, process, and material costs. The labor cost is negligible due to the fact that it does not critically affect the industrial cost.