A Bayesian network model for supporting the formation of PSS design knowledge

Recently, product-service systems (PSS) have drawn the interest of the manufacturing industry. Designing PSS to enhance the value of their core products, manufacturers should assume that their products are their strength or constraint and also derive the service solution logically. However, PSS design knowledge to determine the services suitable for manufacturers’ core products is unclear. As a result, determining a service solution that is compatible with their core products is difficult. This difficulty consequently prevents the manufacturing industry from realising high-quality PSS. To form PSS design knowledge efficiently, this study aims to support the analysis of the complicated and diverse relationships between product characteristics and service contents. Specifically, a Bayesian network model that represents the logical structure between the product characteristics and service contents common among existing PSS cases is constructed through computational learning based on statistical data on PSS cases.