An Autonomous Agent Framework for Feature-Label Extraction from Device Dialogues and Automatic Multi-Dimensional Device Hosting Planning Based on Large Language Models

TL;DR

AirAgent presents an LLM-driven autonomous agent for home air systems, featuring a two-layer architecture with Memory Tag Extraction and Reasoning-Driven Planning to achieve proactive, personalized air management via a semi-streaming output that simultaneously delivers explicit reasoning and device-executable commands. The work introduces a hierarchical memory tagging schema for user demographics, thermal preferences, health conditions, and memory lifecycle, along with a two-dimensional Inference-Model Tag System to fuse environmental data, user profiles, and device status. Empirical results show substantial gains in user experience and accuracy over baselines, with the in-house models achieving up to 94.9% UX and improved latency, demonstrating the potential for health-aware, real-time air optimization in smart homes. The framework enables a scalable, interpretable, and autonomous air stewardship across 25 indoor parameters under dynamic indoor and outdoor conditions, with implications for safer, more comfortable residential environments.

Abstract

With the deep integration of artificial intelligence and smart home technologies, the intelligent transformation of traditional household appliances has become an inevitable trend. This paper presents AirAgent--an LLM-driven autonomous agent framework designed for home air systems. Leveraging a voice-based dialogue interface, AirAgent autonomously and personally manages indoor air quality through comprehensive perception, reasoning, and control. The framework innovatively adopts a two-layer cooperative architecture: Memory-Based Tag Extraction and Reasoning-Driven Planning. First, a dynamic memory tag extraction module continuously updates personalized user profiles. Second, a reasoning-planning model integrates real-time environmental sensor data, user states, and domain-specific prior knowledge (e.g., public health guidelines) to generate context-aware decisions. To support both interpretability and execution, we design a semi-streaming output mechanism that uses special tokens to segment the model's output stream in real time, simultaneously producing human-readable Chain-of-Thought explanations and structured, device-executable control commands. The system handles planning across 25 distinct complex dimensions while satisfying more than 20 customized constraints. As a result, AirAgent endows home air systems with proactive perception, service, and orchestration capabilities, enabling seamless, precise, and personalized air management responsive to dynamic indoor and outdoor conditions. Experimental results demonstrate up to 94.9 percent accuracy and more than 20 percent improvement in user experience metrics compared to competing commercial solutions.

Figures (3)

• Figure 1: Data Construction for Memory Tag Extraction Model
• Figure 2: Reasoning-planning model semi-streaming output
• Figure 3: Method of Inference and Planning Module