Generative AI-enabled case reconstruction: a new approach for product–service systems conceptual design

Digitalization and servitization are accelerating the shift toward Product-Service Systems (PSS), yet PSS conceptual design lacks a systematic approach to achieve innovation and feasibility simultaneously. To address this gap, this study proposes a Theory of Inventive Problem Solving (TRIZ)-based case reconstruction method. Guided by a six-dimensional ecological model, the method transforms historical PSS cases from intact solutions into reconfigurable knowledge units through deconstruction, recombination, and contextualized expression, thereby mitigating the tension between case reuse and innovation. We further integrate TRIZ inventive principles, a domain-specific PSS framework, and large language model (LLM)-based generative reasoning to facilitate knowledge creation in PSS conceptual design while reducing dependence on expert-driven knowledge elicitation. A multi-expert prompting mechanism is also incorporated to balance competing design requirements across multiple dimensions. A medical 3D printing PSS case was used for validation. The results show that the method can generate PSS concepts that are both innovative and feasible, while providing a more structured and reproducible design process. The study advances PSS design theory from case reuse to knowledge creation and offers practical guidance for firms pursuing digital servitization.