Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Agentic Search Engine for Real-Time IoT Data

Abdelrahman Elewah, Khalid Elgazzar

TL;DR

This paper tackles fragmentation and heterogeneity in IoT data by proposing IoT-ASE, a real-time agentic search engine built atop SensorsConnect. It combines IoT-RAG-SE, which embeds service descriptions and performs semantic search over a vector DB, with a GA-RAG workflow of Classifier, Retriever, Generator, and Reviewer to generate accurate, context-aware responses without embedding live IoT data. In a Toronto case study using 500 services and 37,033 place documents, IoT-ASE achieved an intent retrieval accuracy of $92\%$ and produced concise, preference-aligned recommendations, outperforming a baseline Gemini approach in targeted queries. The work demonstrates that lightweight, real-time IoT contexts can support efficient decision-making and scalable service discovery in edge-to-cloud IoT ecosystems.

Abstract

The Internet of Things (IoT) has enabled diverse devices to communicate over the Internet, yet the fragmentation of IoT systems limits seamless data sharing and coordinated management. We have recently introduced SensorsConnect, a unified framework to enable seamless content and sensor data sharing in collaborative IoT systems, inspired by how the World Wide Web (WWW) enabled a shared and accessible space for information among humans. This paper presents the IoT Agentic Search Engine (IoT-ASE), a real-time search engine tailored for IoT environments. IoT-ASE leverages Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) techniques to address the challenge of searching vast, real-time IoT data, enabling it to handle complex queries and deliver accurate, contextually relevant results. We implemented a use-case scenario in Toronto to demonstrate how IoT-ASE can improve service quality recommendations by leveraging real-time IoT data. Our evaluation shows that IoT-ASE achieves a 92\% accuracy in retrieving intent-based services and produces responses that are concise, relevant, and context-aware, outperforming generalized responses from systems like Gemini. These findings highlight the potential IoT-ASE to make real-time IoT data accessible and support effective, real-time decision-making.

Agentic Search Engine for Real-Time IoT Data

TL;DR

This paper tackles fragmentation and heterogeneity in IoT data by proposing IoT-ASE, a real-time agentic search engine built atop SensorsConnect. It combines IoT-RAG-SE, which embeds service descriptions and performs semantic search over a vector DB, with a GA-RAG workflow of Classifier, Retriever, Generator, and Reviewer to generate accurate, context-aware responses without embedding live IoT data. In a Toronto case study using 500 services and 37,033 place documents, IoT-ASE achieved an intent retrieval accuracy of and produced concise, preference-aligned recommendations, outperforming a baseline Gemini approach in targeted queries. The work demonstrates that lightweight, real-time IoT contexts can support efficient decision-making and scalable service discovery in edge-to-cloud IoT ecosystems.

Abstract

The Internet of Things (IoT) has enabled diverse devices to communicate over the Internet, yet the fragmentation of IoT systems limits seamless data sharing and coordinated management. We have recently introduced SensorsConnect, a unified framework to enable seamless content and sensor data sharing in collaborative IoT systems, inspired by how the World Wide Web (WWW) enabled a shared and accessible space for information among humans. This paper presents the IoT Agentic Search Engine (IoT-ASE), a real-time search engine tailored for IoT environments. IoT-ASE leverages Large Language Models (LLMs) and Retrieval Augmented Generation (RAG) techniques to address the challenge of searching vast, real-time IoT data, enabling it to handle complex queries and deliver accurate, contextually relevant results. We implemented a use-case scenario in Toronto to demonstrate how IoT-ASE can improve service quality recommendations by leveraging real-time IoT data. Our evaluation shows that IoT-ASE achieves a 92\% accuracy in retrieving intent-based services and produces responses that are concise, relevant, and context-aware, outperforming generalized responses from systems like Gemini. These findings highlight the potential IoT-ASE to make real-time IoT data accessible and support effective, real-time decision-making.

Paper Structure

This paper contains 28 sections, 14 figures, 1 table.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Background
  4. SensorsConnect Framework
  5. Related Work
  6. Large Language Model
  7. Retrieval-Augmented Generation RAG
  8. Agentic LLM System
  9. Sensors Data Retrieval Based on LLM
  10. System Architecture
  11. Data Model
  12. IoT-Retrieval-Augmented Generation-Search Engine (IoT-RAG-SE)
  13. Services descriptions Embedding
  14. Performing semantic search
  15. Generic Agentic RAG (GA-RAG)
  16. ...and 13 more sections

Figures (14)

  • Figure 1: SensorsConnect Architecture
  • Figure 2: IoT-ASE Logical Data Model
  • Figure 3: IoT Retrieval-Augmented Generation-Search Engine (IoT-RAG-SE)
  • Figure 4: Generic Agentic RAG (GA-RAG) System State Diagram
  • Figure 5: IoT Agentic Search Engine (IoT-ASE)
  • ...and 9 more figures