Reproduce “Internet of Agentic Things: Networked AI Agents for Closed-Loop IoT Orchestration” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Quanyan Zhu

The Internet of Agentic Things (IoAT) framework, developed by Quanyan Zhu at New York University, proposes a three-layered architecture and a formal workflow-control model to enable networked AI agents for closed-loop orchestration of cyber-physical systems. It transitions traditional IoT from reactive monitoring to proactive, adaptive autonomy by integrating capabilities like foundation models and digital twins, addressing the complexity of large-scale physical environments.

Key insights

  • IoAT represents a convergence of IoT, CPS, edge computing, digital twins, and agentic AI, forming a distributed architecture for scalable Physical AI.
  • The hylomorphic dynamic programming model effectively bridges high-level agentic intent ('unfold') with physical actions and feeds physical outcomes ('fold') back to inform strategic planning, ensuring dynamic consistency.
  • Integrating digital twins as active services allows agents to evaluate candidate actions and forecast states, enhancing safety and decision quality before physical execution in critical infrastructures.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.