The role of the fuzzy cognitive map in hierarchical semantic net-based assembly design decision making

For today's highly industrialised and internet-centred business world, it has become essential for manufacturing companies to devise more intelligent assembly design (AsD) and decision support tools to promote customer satisfaction. In order to achieve high performance throughout a product's life-cycle, an AsD system should be able to assist designers' decision-making during assembly and joint design processes in order to avoid design specification violation. An assembly design decision (ADD) problem occurs when the current AsD violates assembly specifications. To obtain a robust design, appropriate joints should be determined by considering the mechanical and mathematical implications of assembly/joining. To tackle the ADD problem, we introduce a hierarchical semantic net (HSN) model to represent evaluation and AsD knowledge, which are of course present, both in the multi-criteria and in a knowledge-based ADD problem. However, the HSN model still requires a methodology for capturing manufacturing environment knowledge and utilising that knowledge for assembly design decision making (ADDM). In the present paper, to avoid the subjective criterion-weight determination of traditional multi-criteria evaluation techniques, we use a fuzzy cognitive map (FCM) in the ADDM framework. An Assembly Advisory (AsA) engine and FCM simulator are developed to implement the presented ADDM framework. We conduct experiments for a realistic welded structure using several scenarios to show the ADDM framework's feasibility. This paper shows that an FCM can be successfully utilised to determine weights of ADDM criteria while capturing manufacturing environment knowledge. An FCM also provides a rich environment for ‘what-if’ experiments to determine ADD criterion weights.