Pith. sign in

REVIEW 3 major objections 6 minor 65 references

Agentic IoT: Architectures, Applications, and Challenges Toward the Internet of Agents

T0 review · 3 major / 6 minor · reviewed 2026-07-11 · grok-4.5

Pith's one-line read Agentic IoT turns connected devices into distributed cognitive agents that perceive, reason, plan, act, and learn across the device-edge-cloud continuum.

desk verdict Clean field-organizing survey that coins Agentic IoT and ships a usable three-tier architecture; no new empirics, and the elastic MCU-to-cloud loop is openly left as open research. read the letter →

arxiv 2607.04219 v1 pith:FFKXLDWL submitted 2026-07-05 cs.AI cs.MAcs.NI

classification cs.AIcs.MAcs.NI
keywords AgenticIoTInternetofAgentsautonomousLargeLanguageModelsedge-fog-cloudcontinuummulti-agentsystemsAIocyber-physical
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Most AIoT systems still run task-specific models that only classify or forecast from sensor data. This paper argues that IoT should instead become ecosystems of autonomous agents that share a common cognitive loop and operate across the device, edge/fog, and cloud layers. It formally defines Agentic IoT as that next-generation paradigm, positions it relative to AIoT and the Internet of Agents, and supplies a three-tier reference architecture with a cross-layer intelligence plane. The architecture distributes perception-action near devices, tactical decisions at the edge, and global planning in the cloud, while shared modules handle memory, tool use, protocol bridging, and safety. A literature review shows the field moving from symbolic multi-agent systems toward LLM-driven agents, and the paper maps the remaining barriers—interoperability, trust, resource limits, real-time reasoning, lifelong memory, coordination, and explainability—into concrete research directions. A sympathetic reader cares because the same stack that today only senses and reacts could, under this design, plan and coordinate physical actions without constant human orchestration.

What carries the argument

The three-tier reference architecture with a cross-layer agentic intelligence plane: device, edge/fog, and cloud agents all implement the same perceive-reason-plan-act-learn loop, while shared modules for memory/RAG, tool use, communication (MCP/A2A plus IoT protocol bridging), coordination, and safety scale intelligence to the resources of each tier.

What would settle it

Deploy a hybrid multi-tier agent system on a real resource-constrained IoT testbed with intermittent links; measure whether local device/edge agents continue correct goal-directed physical actions under the latest policies while cloud agents refine plans, without violating latency, energy, or safety bounds.

Watch

Extended reading notes

Core claim

Agentic IoT is a next-generation cognitive IoT paradigm that transforms networks of connected devices into distributed cognitive agent ecosystems by integrating the perception, reasoning, planning, learning, and action capabilities of autonomous AI agents with cyber-physical systems, operating across the device/edge-fog-cloud continuum under a shared agentic loop and shared support modules.

Load-bearing premise

That one shared cognitive loop and shared support modules can run elastically from microcontrollers to cloud-scale models and still stay real-time, energy-feasible, and safe when connectivity is intermittent.

Editorial extensions

If this is right

  • IoT deployments can shift from reactive sensing and fixed ML pipelines to goal-driven agents that replan under changing conditions.
  • Edge gateways become semantic bridges that convert device telemetry into agent context and agent decisions into device commands.
  • Safety and governance constraints can be enforced uniformly on every autonomous action, from a single actuator to a global policy update.
  • Domain systems in smart cities, disaster response, agriculture, industry, and networks can reuse the same cognitive architecture rather than one-off models.
  • Future work is steered toward foundation agents for IoT, edge-native designs, agentic digital twins, lifelong self-adaptation, process-level explainability, energy-proportional agents, and shared simulation benchmarks.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the shared-loop-plus-shared-modules design holds, many existing AIoT point solutions can be re-hosted as specialized agents under one coordination fabric instead of remaining isolated models.
  • Protocol bridging at the edge may become the practical bottleneck: without a stable semantic boundary between MQTT/CoAP and MCP/A2A, the upper agent layer cannot stay independent of the installed device base.
  • The same architecture that enables physical tool use also multiplies accountability risk; process-level provenance and agentic XAI are not optional add-ons but load-bearing requirements for deployment.
  • Energy-proportional agents that dial capability up or down by task risk would make lifelong on-device learning feasible where always-on large models cannot run.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This survey paper defines Agentic IoT as a next-generation cognitive IoT paradigm that integrates autonomous AI agents (perception, reasoning, planning, learning, action, tool use) with cyber-physical systems, transforming IoT from data-centric sensing into distributed agent ecosystems across the device–edge/fog–cloud continuum. It positions the concept relative to IoT, AIoT, multi-agent systems, and the Internet of Agents (Definition box; Table I; §II), proposes a three-tier reference architecture with a cross-layer agentic intelligence plane, shared cognitive loop, memory/RAG/tool modules, and protocol bridging (Fig. 1; §III), reviews domain literature via comparative tables (Tables II–III; §IV), and catalogs challenges and research directions (§V).

Significance. If the framing holds, the paper supplies a useful organizing vocabulary and reference architecture for an emerging intersection of LLM agents and IoT. Strengths include a clear paradigm comparison (Table I), a concrete three-tier design with protocol bridging and hybrid deployment (§III-D–E), a practical SLM comparison for constrained devices (Table II), and an explicit open-problem agenda (Fig. 2; C1–C7, D1–D8). As a survey-plus-architecture piece it does not claim empirical validation; its value is field-structuring rather than a new theorem or system result. That contribution is real for a community still fragmented across IoA, AIoT, and agentic AI, provided the architecture is treated as a hypothesis rather than a proven stack.

major comments (3)
  1. §III-B and §III-E (and Challenges C3/C4): The architecture’s load-bearing premise is that one shared agentic loop (perceive–reason–plan–act–learn) plus shared modules for memory, tool use, and safety can scale elastically from TinyML/MCU agents to cloud LLMs while remaining real-time, energy-feasible, and safe under intermittent connectivity. The paper correctly lists this as open research, but the body of §III still presents the loop and hybrid deployment as the operational model without any quantitative feasibility bounds, latency budgets, or failure modes for the nested timescales. A short feasibility subsection (even order-of-magnitude latency/energy envelopes or a worked hybrid example) would make the central architectural claim defensible rather than purely stipulative.
  2. §IV and Table III: The literature review is organized by domain and mechanism, which is valuable, but the mapping from reviewed systems to the proposed architecture is mostly descriptive. Several cited works predate LLM agents or use classical BDI/FIPA stacks; the paper does not systematically score how far each study realizes the five-stage loop, cross-layer plane, or protocol bridging of Fig. 1. Without that gap analysis (or an explicit “coverage vs. missing capability” column), the claim that the architecture unifies the field remains under-supported by the evidence presented.
  3. Safety and governance module (§III-A, shared plane; Challenge C2): Placing a single cross-layer safety-and-governance module over every actuator command and tool call is a strong design choice, yet the manuscript gives no concrete enforcement model (policy language, verification point, fail-safe behavior when the uplink is down). Given that tool calls can produce physical effects (§III-C), this is load-bearing for the architecture’s credibility in cyber-physical settings and needs at least a sketch of mechanisms and residual risks.
minor comments (6)
  1. Table I: “Task-spesific” is misspelled; also “spesific” appears elsewhere. Standardize spelling and terminology (Agentic IoT vs. Internet of Agents vs. IoA) on first use in each major section.
  2. Fig. 1: The cross-layer plane and vertical/horizontal flows are dense; a short caption walkthrough of one end-to-end example (e.g., anomaly → edge plan → cloud policy update) would improve readability.
  3. Table II: “Ltd.” tool support and the footnote on SmolLM2 are useful; state the selection criteria and cutoff date for models so the table ages more gracefully.
  4. §IV subsections mix older symbolic IoA work with 2025–2026 LLM-agent papers; a brief chronological or capability-level framing sentence at the start of each domain would reduce the impression of an unweighted catalog.
  5. References: Several arXiv and 2025–2026 conference items are appropriate for a fast-moving area; ensure DOIs/venue details are complete where available and that self-citations are balanced against independent IoA/AIoT surveys.
  6. Declaration on Generative AI is present and clear; keep it, and ensure any AI-assisted phrasing in the architecture sections was checked for technical accuracy (e.g., protocol names MCP/A2A/ACP).

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: conceptual survey with stipulative definition and literature synthesis; no fitted parameters, self-referential derivations, or load-bearing self-citation chains.

full rationale

This is a field-organizing survey paper, not a formal or empirical derivation. The central claim is a stipulative definition of Agentic IoT (Definition box, Abstract, §I–II) that organizes existing AIoT, multi-agent, and Internet-of-Agents literature into a three-tier reference architecture (Fig. 1, §III) with a shared agentic loop and modules. No equations appear; no parameters are fitted to data and then re-presented as predictions; no uniqueness theorems or ansatzes are imported from the authors’ prior work to force the architecture. Self-positioning language (“to the best of our knowledge, there is still no comprehensive survey”) is ordinary survey framing and does not underwrite any technical result. Domain applications (§IV) and challenges (§V) are literature reviews and open-problem statements, not closed-form reductions. The paper is therefore self-contained as a conceptual synthesis; circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 2 invented entities

Conceptual survey; load-bearing premises are domain assumptions about the feasibility of LLM-style agents on constrained hardware and the sufficiency of existing agent protocols. No free parameters or fitted constants. Invented entities are the named paradigm and the architectural plane.

assumptions (4)
  • domain assumption Autonomous AI agents possessing perception, multi-step reasoning (CoT/ReAct), planning, tool use, memory, and multi-agent coordination can be realized with current or near-term LLMs/SLMs.
    Invoked throughout Sections I–III and Table I; underpins the claim that Agentic IoT is achievable rather than purely aspirational.
  • domain assumption Intelligence can be elastically partitioned across device/edge/cloud tiers so that the same agentic loop runs at different depths and timescales without breaking real-time or energy constraints.
    Core of the three-tier architecture and hybrid deployment strategy (Section III-A, III-E).
  • domain assumption Emerging agent protocols (MCP, A2A, ACP) plus edge-level protocol bridging are sufficient to connect heterogeneous IoT devices to agent ecosystems.
    Section III-D; treated as an enabling technology rather than an open research problem that might invalidate the architecture.
  • ad hoc to paper A shared cross-layer safety-and-governance module can enforce operational constraints on every autonomous action from device to cloud.
    Introduced in the agentic intelligence plane (Fig. 1 and Section III-A); not derived from prior standards.
invented entities (2)
  • Agentic IoT (as a named paradigm)
    purpose: To unify AIoT, multi-agent systems, and Internet-of-Agents literature under a single IoT-centric cognitive framework.
    Explicitly defined in the Definition box; claimed as previously unsystematized.
  • Cross-layer Agentic Intelligence Plane
    purpose: Provides a common agentic loop and shared modules (memory, tools, communication, coordination, safety) that all tiers reuse.
    Central architectural novelty of Fig. 1 and Section III-A; not present in conventional layered IoT stacks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Agentic IoT: Architectures, Applications, and Challenges Toward the Internet of Agents." pith.science (2026). https://pith.science/paper/FFKXLDWL

@misc{pith2026260704219,
  author       = {Pith},
  title        = {Pith review of: Agentic IoT: Architectures, Applications, and Challenges Toward the Internet of Agents},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FFKXLDWL}},
  note         = {Machine review of arXiv:2607.04219}
}
read the original abstract

The integration of AI into Internet of Things (AIoT) systems has gradually transformed them from passive data collection infrastructures into intelligent systems capable of anomaly detection, predictive maintenance, classification, forecasting, and optimization. However, most existing solutions still rely on task-specific models that infer from sensor data; thus, system-wide capabilities such as real-time reasoning, adaptive planning, autonomous coordination, learning, tool use, and contextual decision-making remain limited. This paper examines Agentic IoT as a next-generation cognitive IoT paradigm that integrates the perception, reasoning, planning, learning, and action capabilities of autonomous AI agents with cyber-physical systems. Agentic IoT aims to transform IoT from data-centric sensing and inference infrastructures into distributed cognitive agent ecosystems operating across the device/edge-fog-cloud continuum. The paper first grounds this transition as a paradigm shift and positions Agentic IoT in relation to AIoT, edge intelligence, multi-agent systems, and the Internet of Agents. It then systematically reviews current studies, presents a holistic architectural framework, discusses domain-specific application potential, and identifies key technical, operational, and research challenges together with future research directions.

Figures

Figures reproduced from arXiv: 2607.04219 by the authors.

Figure 1
Figure 1. Three-Tier Reference Architecture for Agentic IoT [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Challenges and future research directions in Agentic IoT. [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

65 extracted references · 1 canonical work pages

  1. [1]

    The evolution of AI: From classical machine learning to modern large language models,

    R. Mundlamuri, G. R. Gunnam, N. K. Mysari, and J. Pujuri, “The evolution of AI: From classical machine learning to modern large language models,”IEEE Access, vol. 13, pp. 178 302–178 341, 2025

  2. [2]

    Agentic AI: a com- prehensive survey of architectures, applications, and future directions,

    M. Abou Ali, F. Dornaika, and J. Charafeddine, “Agentic AI: a com- prehensive survey of architectures, applications, and future directions,” Artificial Intelligence Review, vol. 59, no. 1, p. 11, 2025

  3. [3]

    Deep learning based delay and bandwidth efficient data transmission in iot,

    ˙I. K ¨ok, B. H. C ¸ orak, U. Yavano˘glu, and S. ¨Ozdemir, “Deep learning based delay and bandwidth efficient data transmission in iot,” in2019 IEEE International Conference on Big Data (Big Data). IEEE, 2019, pp. 2327–2333

  4. [4]

    Smart x: A description for smart environments,

    R. Dabels, “Smart x: A description for smart environments,” inPro- ceedings of the IADIS International Conference Mobile Learning 2023, P. Kommers, I. A. S ´anchez, and P. Isa´ıas, Eds. IADIS Press, 2023, pp. 103–110

  5. [5]

    Agentic AI: Autonomous intelligence for complex goals—a comprehensive survey,

    D. B. Acharya, K. Kuppan, and B. Divya, “Agentic AI: Autonomous intelligence for complex goals—a comprehensive survey,”IEEE Access, vol. 13, pp. 18 912–18 936, 2025

  6. [6]

    Internet of agents: Fundamentals, applications, and challenges,

    Y . Wang, S. Guo, Y . Pan, Z. Su, F. Chen, T. H. Luan, P. Li, J. Kang, and D. Niyato, “Internet of agents: Fundamentals, applications, and challenges,”IEEE Transactions on Cognitive Communications and Net- working, vol. 12, pp. 4476–4501, 2026

  7. [7]

    The internet of AI agents (iaia): A new frontier in networked and distributed intelligence,

    P. Dazzi, “The internet of AI agents (iaia): A new frontier in networked and distributed intelligence,”International Journal of Networked and Distributed Computing, vol. 13, no. 1, p. 16, 2025

  8. [8]

    Complementing iot services through software defined networking and edge computing: A comprehensive survey,

    W. Rafique, L. Qi, I. Yaqoob, M. Imran, R. U. Rasool, and W. Dou, “Complementing iot services through software defined networking and edge computing: A comprehensive survey,”IEEE Communications Sur- veys & Tutorials, vol. 22, no. 3, pp. 1761–1804, 2020

Show all 65 references
  1. [9]

    Cognitive internet of things: A review of theory, applications, and recent advances,

    A. Giuliano, A. McCafferty-Leroux, J. Yawney, and S. Andrew Gadsden, “Cognitive internet of things: A review of theory, applications, and recent advances,”IEEE Communications Surveys & Tutorials, vol. 28, pp. 446– 484, 2026

  2. [10]

    Internet of agents: Fundamentals, applications, and challenges,

    Y . Wang, S. Guo, Y . Pan, Z. Su, F. Chen, T. H. Luan, P. Li, J. Kang, and D. Niyato, “Internet of agents: Fundamentals, applications, and challenges,”IEEE Transactions on Cognitive Communications and Net- working, 2025

  3. [11]

    AI agents vs. Agentic AI: A conceptual taxonomy, applications and challenges,

    R. Sapkota, K. I. Roumeliotis, and M. Karkee, “AI agents vs. Agentic AI: A conceptual taxonomy, applications and challenges,”Information Fusion, p. 103599, 2025

  4. [12]

    Model-based trustworthiness evaluation of autonomous cyber-physical production systems: A systematic mapping study,

    M. Zahid, A. Bucaioni, and F. Flammini, “Model-based trustworthiness evaluation of autonomous cyber-physical production systems: A systematic mapping study,”ACM Comput. Surv., vol. 56, no. 6, Feb

  5. [13]

    Available: https://doi.org/10.1145/3640314

    [Online]. Available: https://doi.org/10.1145/3640314

  6. [14]

    Autonomous cyber- physical systems enabling smart positive energy districts,

    D. Siakas, G. Lampropoulos, and K. Siakas, “Autonomous cyber- physical systems enabling smart positive energy districts,”Applied Sciences, vol. 15, no. 13, p. 7502, 2025

  7. [15]

    Content-centric data and computation offload- ing in AI-supported fog networks for next generation iot,

    I. Koek and S. ¨Ozdemir, “Content-centric data and computation offload- ing in AI-supported fog networks for next generation iot,”Pervasive and Mobile Computing, vol. 85, p. 101654, 2022

  8. [16]

    Artificial intelligence of things: A survey,

    S. I. Siam, H. Ahn, L. Liu, S. Alam, H. Shen, Z. Cao, N. Shroff, B. Krishnamachari, M. Srivastava, and M. Zhang, “Artificial intelligence of things: A survey,”ACM Transactions on Sensor Networks, vol. 21, no. 1, pp. 1–75, 2025

  9. [17]

    From LLM reasoning to autonomous AI agents: A comprehensive review,

    M. A. Ferrag, N. Tihanyi, and M. Debbah, “From LLM reasoning to autonomous AI agents: A comprehensive review,”IEEE Access, vol. 14, pp. 84 237–84 285, 2026

  10. [18]

    A survey of agent interoperability protocols: Model context protocol (mcp), agent communication protocol (acp), agent-to-agent protocol (a2a), and agent network protocol (anp),

    A. Ehtesham, A. Singh, G. K. Gupta, and S. Kumar, “A survey of agent interoperability protocols: Model context protocol (mcp), agent communication protocol (acp), agent-to-agent protocol (a2a), and agent network protocol (anp),”arXiv preprint arXiv:2505.02279, 2025

  11. [19]

    Acps: Agent collaboration protocols for the internet of agents,

    C. Li, J. Wu, Q. Du, S. Yu, R. Zou, K. Yu, K. Chen, K. Li, Y . Qian, X. Guoet al., “Acps: Agent collaboration protocols for the internet of agents,” in2025 9th IEEE International Conference on Network Intelligence and Digital Content (IC-NIDC). IEEE, 2025, pp. 342– 346

  12. [20]

    From internet of things to internet of agents,

    H. Yu, Z. Shen, and C. Leung, “From internet of things to internet of agents,” in2013 IEEE international conference on green computing and communications and IEEE Internet of Things and IEEE cyber, physical and social computing. IEEE, 2013, pp. 1054–1057

  13. [21]

    Semantic agent contracts for internet of agents,

    P. Pico-Valencia and J. A. Holgado-Terriza, “Semantic agent contracts for internet of agents,” in2016 IEEE/WIC/ACM International Confer- ence on Web Intelligence Workshops (WIW). IEEE, 2016, pp. 76–79

  14. [22]

    Integration of MultiAgent systems with resource-oriented architecture for management of IoT- objects,

    P. Pico Valencia and J. A. Holgado-Terriza, “Integration of MultiAgent systems with resource-oriented architecture for management of IoT- objects,” inIntelligent Environments 2018. IOS Press, 2018, pp. 567– 576

  15. [23]

    Towards the internet of agents: an analysis of the internet of things from the intelligence and autonomy perspective,

    P. Pico-Valencia, J. A. Holgado-Terriza, D. Herrera-S ´anchez, and J. Sampietro, “Towards the internet of agents: an analysis of the internet of things from the intelligence and autonomy perspective,”Ingenier ´ıa e Investigaci´on, vol. 38, no. 1, pp. 121–129, 2018

  16. [24]

    A sys- tematic method for building internet of agents applications based on the linked open data approach,

    P. Pico-Valencia, J. A. Holgado-Terriza, and P. Paderewski, “A sys- tematic method for building internet of agents applications based on the linked open data approach,”Future generation computer systems, vol. 94, pp. 250–271, 2019

  17. [25]

    An internet of agents architecture for training and deployment of deep convolutional models,

    L. Rodriguez-Benitez, C. C. Ruiz, L. C. G ´omez, R. Herv ´as, and L. Jimenez-Linares, “An internet of agents architecture for training and deployment of deep convolutional models,”Journal of Signal Processing Systems, vol. 94, no. 3, pp. 283–291, 2022

  18. [26]

    Acting as a trustee for internet of agents in the absence of explicit feedback,

    A. Aref and T. Tran, “Acting as a trustee for internet of agents in the absence of explicit feedback,” inInternational Conference on E- Technologies. Springer, 2017, pp. 3–23

  19. [27]

    An integrated trust establishment model for the Internet of Agents,

    A. Aref and T. Tran, “An integrated trust establishment model for the Internet of Agents,”Knowledge and Information Systems, vol. 62, no. 1, pp. 79–105, 2020

  20. [28]

    Securing Agentic AI in IoT systems,

    S. Kumi, R. K. Lomotey, and R. Deters, “Securing Agentic AI in IoT systems,” in2025 IEEE International Conference on Smart Internet of Things (SmartIoT). IEEE, 2025, pp. 199–206

  21. [29]

    Toward the internet of Agentic AI: Protocols, architecture, and challenges,

    Y . Ren, J. Yang, H. Zhang, F. R. Yu, Y . Xiao, X. Cao, C. Sun, and Y . He, “Toward the internet of Agentic AI: Protocols, architecture, and challenges,”IEEE Communications Magazine, 2026

  22. [30]

    Agentic intelligence for unified cyber defense: A self- adaptive framework for threat detection across cloud, edge, and iot systems,

    B. Vijetha, “Agentic intelligence for unified cyber defense: A self- adaptive framework for threat detection across cloud, edge, and iot systems,”IEEE Access, vol. 14, pp. 5104–5118, 2026

  23. [31]

    Integration of multiagent systems with resource-oriented architecture for management of iot- objects,

    P. Pico-Valencia and J. A. Holgado-Terriza, “Integration of multiagent systems with resource-oriented architecture for management of iot- objects,” inIntelligent Environments 2018. IOS Press, 2018, pp. 567– 576

  24. [32]

    Towards an internet of agents model based on linked open data approach,

    P. Pico-Valencia, J. A. Holgado-Terriza, and J. A. Senso, “Towards an internet of agents model based on linked open data approach,” Autonomous Agents and Multi-Agent Systems, vol. 33, no. 1, pp. 84– 131, 2019

  25. [33]

    Draco: Decentralized asynchronous federated learning over row-stochastic wireless networks,

    E. Jeong and M. Kountouris, “Draco: Decentralized asynchronous federated learning over row-stochastic wireless networks,”IEEE Open Journal of the Communications Society, 2025

  26. [34]

    Stackelberg game-based Agentic AI task offloading in internet of agents,

    K. Jiang, Y . Zhong, J. Kang, M. Dai, C. Chen, and S. Xie, “Stackelberg game-based Agentic AI task offloading in internet of agents,” in2025 IEEE 28th International Conference on Computational Science and Engineering (CSE). IEEE, 2025, pp. 9–15

  27. [35]

    Agentic search engine for real- time internet of things data,

    A. Elewah, K. Elgazzar, and S. Elnaffar, “Agentic search engine for real- time internet of things data,”Sensors, vol. 25, no. 19, p. 5995, 2025

  28. [36]

    Multimodal and Agentic intelligence–driven ml fusion for sustainable 6g network slicing,

    S. Vashisht, S. Rani, and H. Feng, “Multimodal and Agentic intelligence–driven ml fusion for sustainable 6g network slicing,”IEEE Open Journal of the Communications Society, 2026

  29. [37]

    Internet of agents framework for connected vehicles: A case study on distributed traffic control system,

    K.-H. N. Bui and J. J. Jung, “Internet of agents framework for connected vehicles: A case study on distributed traffic control system,”Journal of Parallel and Distributed Computing, vol. 116, pp. 89–95, 2018

  30. [38]

    A conceptual framework for smart disaster response using ioa and iot for smart cities,

    H. Kamel, “A conceptual framework for smart disaster response using ioa and iot for smart cities,” in2025 3rd International Conference on Artificial Intelligence, Blockchain, and Internet of Things (AIBThings). IEEE, 2025, pp. 1–10

  31. [39]

    Multi-agent smart water distribution with computer vision-based container detection and AI-driven behavior forecasting for disaster relief in smart cities,

    W. Chumyen, “Multi-agent smart water distribution with computer vision-based container detection and AI-driven behavior forecasting for disaster relief in smart cities,” in2025 IEEE International Conference on Agentic AI (ICA). IEEE, 2025, pp. 7–11

  32. [40]

    Edge-enabled smart agriculture framework: Integrating iot, lightweight deep learning, and Agentic AI for context-aware farming,

    M. U. Tariq, S. M. Saqib, T. Mazhar, M. A. Khan, T. Shahzad, and H. Hamam, “Edge-enabled smart agriculture framework: Integrating iot, lightweight deep learning, and Agentic AI for context-aware farming,” Results in Engineering, p. 107342, 2025

  33. [41]

    Cottonbot: An AI-driven cotton farming assistant and irrigation advisor using LLM-rag and Agentic AI tools,

    D. F. Kandamali, W. M. Porter, E. Porter, A. McLemore, and G. C. Rains, “Cottonbot: An AI-driven cotton farming assistant and irrigation advisor using LLM-rag and Agentic AI tools,”Smart Agricultural Technology, p. 101640, 2025

  34. [42]

    Smart agents as cus- tomers: A semi-autonomous digital commerce model for industry 5.0,

    M. S ¸. S¨umer, T. ¨Ozsoy, F. Y . Okay, and ˙I. K ¨ok, “Smart agents as cus- tomers: A semi-autonomous digital commerce model for industry 5.0,” in2025 7th International Congress on Human-Computer Interaction, Optimization and Robotic Applications (ICHORA). IEEE, 2025, pp. 1–7

  35. [43]

    LLM-driven approach to automated sustainability of iot systems,

    N. Petrovi ´c, D. Krsti´c, S. Suljovi´c, and D. Javor, “LLM-driven approach to automated sustainability of iot systems,” in2025 IEEE 34th Interna- tional Conference on Microelectronics (MIEL). IEEE, 2025, pp. 1–4

  36. [44]

    Harmony: A framework for multimodal LLM-powered AI agents in smart homes via the model context protocol,

    K. Mohanaprasad, G. Rohith, A. Mondal, S. Singh, S. Tiwari, and J. Sharma, “Harmony: A framework for multimodal LLM-powered AI agents in smart homes via the model context protocol,”IEEE Access, 2026

  37. [45]

    Securing Agentic AI in iot systems,

    S. Kumi, R. K. Lomotey, and R. Deters, “Securing Agentic AI in iot systems,” in2025 IEEE International Conference on Smart Internet of Things (SmartIoT), 2025, pp. 199–206

  38. [46]

    Secure multi-LLM Agentic AI and agentification for edge general intelligence by zero-trust: A survey,

    Y . Liu, R. Zhang, H. Luo, Y . Lin, G. Sun, D. Niyato, H. Du, Z. Xiong, Y . Wen, A. Jamalipouret al., “Secure multi-LLM Agentic AI and agentification for edge general intelligence by zero-trust: A survey,” arXiv preprint arXiv:2508.19870, 2025

  39. [47]

    Small language models are the future of Agentic AI,

    P. Belcak, G. Heinrich, S. Diao, Y . Fu, X. Dong, S. Muralidharan, Y . C. Lin, and P. Molchanov, “Small language models are the future of Agentic AI,”arXiv preprint arXiv:2506.02153, 2025

  40. [48]

    Small language models: A systematic review of computational trade-offs, privacy advantages and deployment in intelligent systems,

    S. Dikici and T. T. Bilgin, “Small language models: A systematic review of computational trade-offs, privacy advantages and deployment in intelligent systems,”Expert Systems, vol. 43, no. 8, p. e70331, 2026

  41. [49]

    Minimizing response latency in llm-based agent systems: A comprehensive survey,

    G. Park, S. Lee, and Y . Park, “Minimizing response latency in llm-based agent systems: A comprehensive survey,”IEEE Access, 2026

  42. [50]

    Cognitive edge computing: A compre- hensive survey on optimizing large models and AI agents for pervasive deployment,

    X. Wang, Q. Li, and W. Jia, “Cognitive edge computing: A compre- hensive survey on optimizing large models and AI agents for pervasive deployment,”arXiv preprint arXiv:2501.03265, 2025

  43. [51]

    Reliability in internet of things: Current status and future perspectives,

    L. Xing, “Reliability in internet of things: Current status and future perspectives,”IEEE Internet of Things Journal, vol. 7, no. 8, pp. 6704– 6721, 2020

  44. [52]

    Multi-agent collaboration mechanisms: A survey of LLMs,

    K.-T. Tran, D. Dao, M.-D. Nguyen, Q.-V . Pham, B. O’Sullivan, and H. D. Nguyen, “Multi-agent collaboration mechanisms: A survey of LLMs,”arXiv preprint arXiv:2501.06322, 2025

  45. [53]

    When IoT meet LLMs: Appli- cations and challenges,

    I. K ¨ok, O. Demirci, and S. ¨Ozdemir, “When IoT meet LLMs: Appli- cations and challenges,” in2024 IEEE International Conference on Big Data (BigData), 2024, pp. 7075–7084

  46. [54]

    Interpreting Agentic systems: Beyond model explanations to system-level accountability,

    J. Zhu, D. Gandhi, H. Joshi, A. R. Mianroodi, S. A. Kocak, and D. Ra- machandran, “Interpreting Agentic systems: Beyond model explanations to system-level accountability,”arXiv preprint arXiv:2601.17168, 2026

  47. [55]

    Deploying foundation model powered agent services: A survey,

    W. Xu, J. Chen, P. Zheng, X. Yi, T. Tian, W. Zhu, Q. Wan, H. Wang, Y . Fan, Q. Suet al., “Deploying foundation model powered agent services: A survey,”IEEE Communications Surveys & Tutorials, 2025

  48. [56]

    A survey of foundation models for iot: Taxonomy and criteria-based analysis,

    H. Wei, D. Y . Lee, S. Rohal, Z. Hu, R. Rossi, S. Fang, and S. Pan, “A survey of foundation models for iot: Taxonomy and criteria-based analysis,”arXiv preprint arXiv:2506.12263, 2025

  49. [57]

    Toward edge general intelligence with Agentic AI and agentification: Concepts, technologies, and future directions,

    R. Zhang, G. Liu, Y . Liu, C. Zhao, J. Wang, Y . Xu, D. Niyato, J. Kang, Y . Li, S. Maoet al., “Toward edge general intelligence with Agentic AI and agentification: Concepts, technologies, and future directions,”IEEE Communications Surveys & Tutorials, vol. 28, pp. 4285–4318, 2026

  50. [58]

    From large AI models to Agentic AI: A tutorial on future intelligent communications,

    F. Jiang, C. Pan, K. Wang, P. Michiardi, O. A. Dobre, and M. Debbah, “From large AI models to Agentic AI: A tutorial on future intelligent communications,”IEEE Journal on Selected Areas in Communications, 2026

  51. [59]

    Understanding world or predicting future? a comprehensive survey of world models,

    J. Ding, Y . Zhang, Y . Shang, Y . Zhang, Z. Zong, J. Feng, Y . Yuan, H. Su, N. Li, N. Sukienniket al., “Understanding world or predicting future? a comprehensive survey of world models,”ACM Computing Surveys, vol. 58, no. 3, pp. 1–38, 2025

  52. [60]

    Agentic digital twins: A taxonomy of capabilities for understanding possible futures,

    C. Burr, M. Enzer, J. Shepherd, and D. Wagg, “Agentic digital twins: A taxonomy of capabilities for understanding possible futures,”arXiv preprint arXiv:2601.18799, 2026

  53. [61]

    V2llm: A digital-twin-aware retrieval-augmented large language model for resilient telemetry recovery and fault diagnosis in v2x,

    B. Hazarika, K. Singh, N.-S. V o, B. Canberk, S. L. Cotton, H. Shin, and T. Q. Duong, “V2llm: A digital-twin-aware retrieval-augmented large language model for resilient telemetry recovery and fault diagnosis in v2x,”IEEE Transactions on Network Science and Engineering, vol. 1...

  54. [62]

    From agentification to self-evolving Agentic AI for wireless networks: Concepts, approaches, and future research direc- tions,

    C. Zhao, R. Zhang, J. Wang, D. Niyato, G. Sun, X. Wang, S. Mao, and A. Jamalipour, “From agentification to self-evolving Agentic AI for wireless networks: Concepts, approaches, and future research direc- tions,”IEEE Communications Magazine, 2026

  55. [63]

    Explainable artificial intelligence (XAI) for internet of things: A survey,

    I. K ¨ok, F. Y . Okay, O. Muyanlı, and S. Ozdemir, “Explainable artificial intelligence (XAI) for internet of things: A survey,”IEEE Internet of Things Journal, vol. 10, no. 16, pp. 14 764–14 779, 2023

  56. [64]

    Toward sustainable Agentic AI systems: A survey of architectures and methodologies,

    Y . Lee and E. Park, “Toward sustainable Agentic AI systems: A survey of architectures and methodologies,”Sustainable Development, 2026

  57. [65]

    LLM and AI agents for autonomous systems: A survey of applications, datasets, and security challenges,

    M. A. Ferrag, A. Lakas, N. Tihanyi, and M. Debbah, “LLM and AI agents for autonomous systems: A survey of applications, datasets, and security challenges,”IEEE Open Journal of Intelligent Transportation Systems, vol. 7, pp. 615–657, 2026

Pith tools

Reviewed July 11, 2026 · model on record in the stance chip above.