Modelling and application of hierarchical joint optimisation for modular product family and supply chain architecture

Modular product family architecture (PFA), in coordination with the supply chain, assists manufacturers in achieving lower costs and higher efficiency by sharing a common platform. Despite its advantages, however, the prevailing practice of PFA emphasises architectural aspects that do not focus on the interface requirements for an efficient supply chain. In particular, the individual modules and components are assumed to have equal and/or fixed connection values, thereby overlooking the impact of modularity on the supply chain architecture (SCA). Explicit considerations of alternative modular configurations can invoke changes in granularity to reduce supply chain costs. Furthermore, the general approach of SCA is predominately focused on cost without emphasising commonality, reducing the benefit of modularity. The major challenge is, therefore, to determine the optimal granularity of modules under a coherent framework of product family modularity and supply chain modularity, which are often widely different. To resolve the problem, a bi-level programming (BLP) model is proposed in which the integrated effects of commonality and cost of supply chain modularity are investigated with the architectural and interface modularity (AIM) of product design. The proposed leader–follower decision structure interactively and hierarchically optimises commonality and cost to ensure product design and supply chain coherence and integrity. The experimental results for refrigerator PFA and its supply chain reveal that the proposed integrated modularisation model saves 11.54% and 5.95% SCA cost compared with cost-based and commonality-based models. In algorithmic comparisons, the proposed nested bi-level particle swarm optimisation (NBL-PSO) shows an enhanced performance for various problem instances with lower standard deviation (design cost: 3.3% and SCA cost: 6.4%) compared with the genetic algorithm.