IEEE Annual Congress on Artificial Intelligence of Things (IEEE AIoT)

14-16 Dec 2026 · Kobe, Japan.

http://www.ieee-aiot.org/2026

Track 1: Agentic AI and IoT

Track Chairs:
Muhammad Bilal, Lancaster University, m.bilal@ieee.org
Yifan Guo, Towson University, yguo@towson.edu
Yuichi Sei, University of Electro-Communications (UEC Tokyo), seiuny@uec.ac.jp

Description

Agentic artificial intelligence (AI) marks a shift from reactive, prompt-driven models to autonomous, goal-directed systems capable of perceiving their environment, planning multi-step actions, invoking tools, collaborating with other agents, and adapting their behavior over time. When this paradigm meets the Internet of Things (IoT), it gives rise to a new class of intelligent cyber-physical systems in which distributed agents — running on devices, at the edge, and in the cloud — sense, reason, negotiate, and act on the physical world with minimal human intervention.

This track invites original contributions that explore the deep integration of agentic AI and IoT. We welcome original works in theoretical foundations, system architectures, communication and networking protocols, learning algorithms, orchestration frameworks, safety and governance mechanisms, and real-world deployments in this emerging area. Of particular interest are works addressing how Large Language Model (LLM)-based agents, multi-agent systems, and tool-using AI can be deployed on resource-constrained IoT devices and across the edge–fog–cloud continuum; how agents coordinate over heterogeneous wireless networks; and how trust, security, and human oversight can be ensured when autonomous agents take consequential actions in the physical world. The track aims to bring together researchers and practitioners from AI, IoT, networking, embedded systems, and human-computer interaction (HCI) to shape the future of agent-driven AIoT.

Track Topics

• Foundations and architectures of agentic AI for IoT systems
• LLM-based agents and multimodal foundation models for IoT
• Multi-agent systems, coordination, negotiation and collaboration in IoT environments
• Tool use, function calling and agent orchestration for IoT devices and services
• Planning, reasoning, and decision-making for embodied AIoT agents
• Perception-reasoning-action loop in agentic AIoT systems
• Memory, reflection, and long-horizon autonomy in IoT agents
• On-device and edge deployment of agentic AI
• Communication protocols and semantic communication for agent-to-agent (A2A) interaction in IoT
• Agentic AI for integrated sensing, communication, and control
• Agentic AI for network management, self-organization, and self-healing in IoT
• Digital twins and agentic AI for real-time IoT orchestration
• Human-agent collaboration and human-in-the-loop AIoT systems
• Safety, alignment, and guardrails for autonomous agents acting on the physical world
• Trust, accountability, and explainability of agentic AIoT systems
• Security and privacy threats specific to agentic AI in IoT
• Verification, validation, and benchmarking of agentic AIoT systems
• Energy-efficient and sustainable agentic AI for IoT
• Federated and continual learning for distributed IoT agents
• Agentic AI for autonomous vehicles, drones, and robotics in IoT contexts
• Agentic AI for smart cities, smart manufacturing (Industry 5.0), and smart healthcare
• Standards, governance, and regulatory considerations for agentic AIoT
• Datasets, simulators, and testbeds for agentic AIoT research
• Case studies and real-world deployments of agentic AI in IoT

Paper Submission and Publication

Details of paper submission and publication can be found here.