AdaHome: An Adaptive Smart Home Assistant using Local Small Language Models
2026-07-20 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors developed AdaHome, a smart home assistant that runs lightweight language models locally, which makes it faster and better for privacy than cloud-based systems. AdaHome uses a smart method to decide whether a command needs simple or more detailed understanding, improving efficiency. It also learns user preferences over time without needing complicated updates. Tests showed AdaHome is more accurate and quicker on clear commands and keeps steady performance on unclear ones, while better adapting to user preferences in ongoing interactions.
Smart home assistantLarge Language Models (LLMs)Local deploymentIntent-aware planningChain-of-DraftPreference adaptationPrompt-based learningModel personalizationLatencyMulti-turn dialogue
Authors
Eu Jin Lim, Zhaoxing Li, Sebastian Stein
Abstract
Smart home assistants interpret a wide range of user commands, from explicit device control to underspecified and preference dependent requests. While recent systems based on Large Language Models (LLMs) improve this capability, they often rely on heavyweight reasoning pipelines and cloud-based deployment, limiting their efficiency and suitability for resource-constrained environments, and raising privacy concerns. In addition, existing approaches provide limited support for stable long-term personalization. To address these issues, we present AdaHome, an adaptive smart home assistant designed for locally deployed small language models in smart home environments. Rather than applying complex reasoning uniformly, AdaHome introduces an intent-aware planning framework that dynamically routes commands either to straightforward prompt-based or lightweight reasoning-based components. For commands requiring interpretation, we adopt a Chain-of-Draft strategy to enable efficient and stable decision-making. To support personalization, we further propose a preference adaptation mechanism that learns from user feedback over time without requiring prompt augmentation or model retraining. We evaluate AdaHome against representative LLM-based baselines under a unified small model setting. AdaHome achieves substantially higher accuracy on direct commands (86.7%) while reducing latency by up to 3$\times$. Furthermore, it maintains competitive performance on ambiguous inputs with lower computational cost. In multi-turn scenarios, AdaHome achieves 88% preference consistency, compared to 52.5% for a prompt augmentation baseline.