Governing intelligent manufacturing innovation ecosystems: adaptive bottlenecks and governance rights configurations

Ecosystem governance research has not fully explained how focal manufacturers govern intelligent innovation when implementation depends on distributed and autonomous ecosystem actors. Drawing on an abductive comparative case study of two focal firm-led intelligent manufacturing ecosystems, this study develops the concept of adaptive bottlenecks, defined as dominant implementation constraints that prevent intelligent functions from being validated, adjusted, or embedded in use. The findings show that focal manufacturers align control rights over structural conditions and decision rights over operational discretion with the dominant bottleneck. Under institutional and cognitive misalignment, centralised architectural control and algorithmic decision convergence reduce interpretive, procedural, and institutional variation, supporting systemic reliable value through greater reviewability, reproducibility, traceability, and institutional legitimacy. Under capability rigidity arising from heterogeneous long-tail field demand, open enabling control and embodied empowerment retain core platform boundaries while distributing situated decision rights to partners, field actors, users, and intelligent systems, supporting generative adaptive value through local adaptation and context-specific solution generation. Together, these configurations form two ideal-typical pathways of bottleneck–governance alignment, termed the prescriptive and symbiotic governance pathways. The study develops a bottleneck-contingent account of ecosystem governance that explains how selective configurations of control and decision rights make intelligent innovation workable across ecosystem boundaries.