A Multi-Domain AI Analytics Framework for Smart Manufacturing Across the Product Lifecycle

Abstract

Smart manufacturing (SM) has emerged as the operational expression of Industry 4.0 and of the broader convergence of artificial intelligence (AI), the industrial internet of things (IIoT), cyber–physical systems (CPS) and digital twins (DTs). While the cumulative number of SM publications has grown roughly eighty-fold over the last decade, most contributions remain locally scoped: a single algorithm in a single workshop, a single sensor family on a single asset class, or a single management process in a single enterprise. The field lacks a view that places those contributions inside a common product-lifecycle (PLM) architecture and that identifies, for each lifecycle stage, which analytics tools are mature, which are experimental, and which are missing. This article develops such a view. We conduct a systematic review of 214 peer-reviewed works published between 2012 and 2025, organise their AI contributions into four interlocking domains—strategy and organisation, value-chain intelligence, management support, and infrastructure and capabilities—and map the resulting topics onto a five-stage PLM backbone. Using the coded corpus we report a maturity radar, a barrier frequency analysis and a topic-share breakdown. We then translate the diagnostic into prescriptive guidance in the form of a five-layer AI-enabled transformation roadmap that covers foundations of AI capability, enterprise-wide integration, process optimisation, system reconfiguration and enterprise innovation. Two cross-cutting findings emerge. First, the dominant weakness of current SM systems is not algorithmic but architectural: data fragmentation, missing standards and poor horizontal–vertical–end-to-end integration explain most of the gap between pilot-scale success and enterprise-scale deployment. Second, the sustainability of SM will be decided less by the raw capability of frontier AI models than by how disciplined enterprises are in building knowledge-management (KM) capability, governance of data and workforce AI literacy. The framework is offered as a diagnostic grid for practitioners and as a structured research agenda for scholars.

Keywords: smart manufacturing, Industry 4.0, artificial intelligence, product lifecycle management, industrial internet of things, digital twin, systematic literature review, enterprise transformation.