Industrial platforms have evolved into meta-organizations—coordinating networks of firms, AI agents, and physical devices around a programmable core. In these settings, everyday micro-decisions increasingly shift to algorithms and edge devices, while humans supervise, interpret, and intervene on exceptions. This raises hard questions: what to automate vs. keep interruptible, how far to decentralize decisions to the edge without losing accountability, how to pair model performance with operator trust, and how to update AI safely in live operations.

HIO addresses these challenges with a unified framework connecting (i) governance-by-design (who decides what, and how that is encoded), (ii) division of labor (sensing → inference → actuation → oversight), and (iii) epistemic alignment (how human and machine knowledge is produced, checked, and shared). We study three complementary sites—autonomous drones, a cross-domain AI integrator, and an enterprise platform backbone—to derive general design principles, metrics, and reference architectures for responsible, high-performing AI in industrial ecosystems. 

Why now?

Platforms increasingly “encode governance” (rules, workflows, escalations) as software and ML pipelines.
Edge autonomy grows; accountability and provenance must keep pace.
High-performance models require mechanisms operators can trust and override.
AI updates must balance reliability with continual learning in safety-critical contexts.

See more at Springer, V., Randhawa, K., Jovanović, M., Ritala, P., & Piller, F. T. (2025). Platform design and governance in industrial markets: Charting the meta-organizational logic. Research Policy, 54(6), 105236. https://doi.org/10.1016/j.respol.2025.105236