Adaptability evaluation of conceptual design for smart product-service system: An integrated rough DEMATEL and Bayesian network model

Smart product service system (PSS) design evaluation plays a crucial role in the early design stage of Smart PSS development. It aims at considering the user's value proposition for Smart PSS in the dimensions of intelligence, sustainability, and society, while also ranking Smart PSS to support delivery risk management under uncertainty. However, related Smart PSS evaluation models advocate the use of multi-criteria decision models for obtaining the intrinsic value of PSS by ignoring the influence of the causality of criteria on Smart PSS design, which results in an unstable evaluation process. Additionally, the adaptiveness of Smart PSS has not been thoroughly explored, and determining the adoption intentions of Smart PSS can be challenging. To fill these gaps, a Smart PSS evaluation model that integrates a rough decision-making trial and evaluation laboratory (DEMATEL) and a Bayesian network (BN) is proposed to support probabilistic reasoning for Smart PSS adaptiveness and reduce the risk of Smart PSS development. First, a set of 17 evaluation criteria is identified, and their causal relationships and weights of the criteria were identified using a rough DEMATEL under uncertainty. Second, a BN-based reasoning model for Smart PSS criteria is developed to evaluate the adaptability of smart PSS by constructing a conditional probability calculation model based on criterion weights. Finally, a multifunctional smart cooking center (MFSCC) case study is presented to demonstrate the applicability of the proposed approach. The belief propagation analysis results revealed the key criteria affecting the MFSCC service system development and provided design strategies to improve the adaptiveness of the Smart PSS.

Automated summary

Smart PSS design evaluation is essential in early design stages, but current models neglect the causal relationships between criteria, leading to unstable evaluation processes. To address this, a DEMATEL and BN integrated evaluation model is proposed, enabling probabilistic reasoning and risk reduction through adaptiveness evaluation. The model identifies causal relationships among 17 criteria, constructs a BN-based reasoning model, and applies it to a case study to improve the adaptiveness of a Smart PSS.