Scholarlink

<p>Edge–Cloud collaborative computing has emerged as a key paradigm for enabling real-time robotic question answering (Q&amp;A) in Internet of Things (IoT) environments by combining local intelligence at the edge with global knowledge processing in the cloud. However, existing approaches often suffer from high communication latency, limited reasoning capability at edge devices, and redundant data transmission caused by centralized processing, which adversely affect real-time responsiveness, scalability, and semantic accuracy. To address these challenges, this paper proposes the Edge–Cloud Knowledge Graph Fusion (ECKGF) framework, which integrates context-aware edge-level inference with cloud-based semantic enrichment through a fusion-oriented architecture. In ECKGF, lightweight knowledge graph modules deployed at the edge enable low-latency reasoning over local sensor data, while the cloud layer performs large-scale knowledge expansion, semantic alignment, and global consistency management. An adaptive synchronization mechanism selectively exchanges semantically relevant knowledge between edge and cloud, reducing communication overhead while preserving reasoning accuracy. The framework is evaluated using an industrial IoT–based smart robotic assistant scenario. Experimental results demonstrate that ECKGF significantly reduces query response time while improving reasoning precision and transmission efficiency, validating its effectiveness for intelligent, low-latency robotic Q&amp;A in IoT networks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Edge–cloud collaborative knowledge graph computing for real-time robot Q&A in IoT networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月22日

查看全文
  • Xiaoyun Chen,
  • Hesen Wang,
  • Xi Yang,
  • Hua Huang,
  • Cheng Wei

摘要

Edge–Cloud collaborative computing has emerged as a key paradigm for enabling real-time robotic question answering (Q&A) in Internet of Things (IoT) environments by combining local intelligence at the edge with global knowledge processing in the cloud. However, existing approaches often suffer from high communication latency, limited reasoning capability at edge devices, and redundant data transmission caused by centralized processing, which adversely affect real-time responsiveness, scalability, and semantic accuracy. To address these challenges, this paper proposes the Edge–Cloud Knowledge Graph Fusion (ECKGF) framework, which integrates context-aware edge-level inference with cloud-based semantic enrichment through a fusion-oriented architecture. In ECKGF, lightweight knowledge graph modules deployed at the edge enable low-latency reasoning over local sensor data, while the cloud layer performs large-scale knowledge expansion, semantic alignment, and global consistency management. An adaptive synchronization mechanism selectively exchanges semantically relevant knowledge between edge and cloud, reducing communication overhead while preserving reasoning accuracy. The framework is evaluated using an industrial IoT–based smart robotic assistant scenario. Experimental results demonstrate that ECKGF significantly reduces query response time while improving reasoning precision and transmission efficiency, validating its effectiveness for intelligent, low-latency robotic Q&A in IoT networks.