Towards Effective Multi-platforming Design of Product Family using Genetic Algorithm

Platform-based product family design is recognized as an effective method to construct a product line that satisfies diverse performance requirements while aiming to keep design and production costs low. The success of the resulting product family often relies on properly resolving the tradeoff between increasing commonality across the family and performance loss compared to individual design. In this paper, a systematic approach is proposed to design the scale-based product family with multi-platforming configuration and it contributes in three aspects. Firstly, the effect of commonality on the related product life-cycle activities is evaluated in the platform decision to decide the expected degree of sharing for each design variable and thus generate the anticipated platform configuration. Secondly, unlike many existing methods that assumes a given single-platform, the proposed method addresses the multi-platforming configuration across the family, and can generate alternative product family settings with different levels of commonality. Finally, the product family design is formulated as a multi-optimization problem and solved using a modified genetic algorithm. An industrial example of a planetary gear train for cordless drills is presented to demonstrate the proposed method.