Aims & Scope

Edge Intelligence and Systems (EIS) is a peer-reviewed, open-access international journal dedicated to disseminating pioneering research and innovative practices at the intersection of artificial intelligence (AI), edge computing, and their integration into system design. The journal champions research that unleashes the full potential of modern and sophisticated AI in pervasive edge scenarios, driven by transformative research on algorithms, software, hardware, networking, architectures, systems, and applications that define the future of edge-native intelligent, interactive, and autonomous technologies. EIS is published quarterly online by Scilight Press.

The scope of EIS encompasses the entire technology stack for edge intelligence and systems, from foundational intelligence and computing paradigms to high-level multidisciplinary applications. We welcome original research papers, review articles, and short communications on the following (but not limited to) topics:

Edge-Native AI

  • Large Language Models and Generative AI at the Edge: Efficient inference, on-device fine-tuning/retraining, parameter-efficient adaptation, and retrieval-augmented generation (RAG) for edge applications.
  • Embodied and Agentic AI: Frameworks for autonomous agents, real-time perception-action cycles, multi-modal sensor fusion, and human-robot interaction on edge hardware.
  • Decentralized and Collaborative AI: Advanced federated learning, swarm intelligence, multi-agent systems, and collaborative perception and decision-making among edge devices.
  • Efficient AI for the Edge: Novel techniques in model compression, quantization, neural architecture search (NAS), and continual learning tailored for resource-constrained systems.

Systems and Hardware for Edge AI

  • AI-centric System Architectures: Novel frameworks for seamless device-edge-cloud collaboration, designed specifically for distributed AI workloads.
  • Hardware Acceleration and Co-design: Neuromorphic computing, domain-specific AI accelerators, and software-hardware co-design for high-performance, low-power edge AI.
  • Resource Orchestration for AI Workloads: Dynamic management, scheduling, and allocation of computation, memory, and network resources for complex AI pipelines at the edge.

Networking for Distributed Intelligence

  • AI-driven Networking: Semantic communication, intelligent routing, and resource allocation for distributed inference and training.
  • Low-latency Communication: Protocols and architectures (including 5G/6G) designed to support real-time control loops for embodied AI and interactive applications.

Applications and Advanced Use Cases

  • Interactive Robotics and Autonomous Systems: Drones, vehicles, and robots capable of complex task execution and collaboration in dynamic environments.
  • On-Device Assistants and Human-AI Interaction: Next-generation smart devices that leverage on-device LLMs for personalized, private, and responsive user experiences.
  • Industrial Metaverse and Digital Twins: Real-time synchronization between physical systems and their virtual counterparts, powered by edge AI.

Trustworthy and Responsible Edge AI

  • Security and Privacy: Novel cryptographic and privacy-preserving techniques for decentralized AI models and data.
  • Ethics and Governance: Frameworks for explainable AI (XAI), fairness, and accountability for autonomous agents operating in the real world.
  • Robustness and Reliability: Ensuring the safety and predictability of edge AI systems in mission-critical applications.