Abstract Manufacturing-as-a-Service (MaaS) facilitates on-demand access to distributed manufacturing resources and services provided by third parties through platform-based ecosystems. It enables a distributed platform service-oriented production, but despite advances in AI-enabled planning, intelligent matchmaking, optimisation-driven orchestration and cyber-physical integration, the literature remains fragmented. Most studies report isolated decision layers (process planning, supplier selection, scheduling, pricing or trust) without proposing lifecycle-integrated architectures. A PRISMA 2020-compliant systematic review was conducted on 62 MaaS studies published between 2013 and 2026 to examine how intelligent manufacturing platforms are implemented and which decision functions they support. This paper analyses the literature identifying: (i) an archetype classification of MaaS solutions, from conceptual frameworks to trust infrastructures; (ii) a granularity-based input–output taxonomy that maps information units to intelligent decisions; and (iii) a limitation analysis highlighting barriers in data readiness, interoperability, governance and real-time orchestration under uncertainty. Findings show persistent silos across engineering intelligence, marketplace design, operational optimisation and governance mechanisms. Building on this evidence, the Company-as-a-Service concept is proposed as an extension that integrates MaaS modules into lifecycle-consistent organisational capabilities, coordinated through a “Digital Thread and Governance Kernel”. This paper designs principles for scalable, governance-aware, AI-enabled manufacturing ecosystems. Implications are discussed for platform developers and manufacturers seeking resilient, compliant deployment across SMEs, multi-site networks, and heterogeneous equipment fleets worldwide.
Pietrangeli et al. (Sun,) studied this question.