Autonomous AI Agents for Edge Intelligence and IoT Systems
Session Overview
Rapid advancements in IoT devices have increased demand for intelligent and autonomous decision making at the edge network. Traditional cloud-based AI systems have become a bottleneck due to modern latency and privacy requirements, especially in IoT deployments. Autonomous AI agents can address these challenges through reasoning, planning, and context-aware capabilities.
However, deployment of AI agents on resource-constrained hardware remains an open and largely unexplored challenge. Key barriers include computational cost, inference latency, multi-agent coordination, fairness, and privacy, especially at large and complex scale.
This special session addresses these challenges through artificial intelligence, autonomous agentic architecture, edge-deployed IoT, and blockchain-enabled decentralized agent coordination. It brings together researchers working on neural network compression, multi-agent coordination, fairness-aware deployment, and cross-domain agentic applications to define the roadmap for future AI systems design and deployment.
Topics of Interest
- Agentic Architectures for Edge and IoT
- Neural Network Compression for Agentic Models
- Multi-Agent Coordination and Management
- Federated and Decentralized Agent Learning
- Fairness, Safety, and Accountability of Deployed Agents
- Explainability and Interpretability of Agent Decisions in IoT Applications
- Blockchain-Based Audit Trails of Agentic Systems
- IoT-Based Applications of Autonomous Agents
- Security, Privacy, and Trust of AI Agents
- Governance of Large and Complex AI Agents in a Smart City Environment
Review and Publication
All submitted papers will undergo peer review by at least two expert reviewers. Acceptance decisions will be based on technical quality, novelty, and relevance to the session's scope of edge AI agents and IoT deployment. Accepted papers will be submitted for inclusion in IEEE Xplore based on the standard AIBThings 2026 procedure.