This project explores how small language models can support intelligent IoT systems directly at the edge. The goal is to enable local reasoning, adaptive automation, and context-aware decision-making while reducing dependence on cloud-scale models for privacy-sensitive smart environments.

Research Focus

• Edge-deployable small language models for intelligent IoT agents
• Local reasoning and automation for privacy-sensitive environments
• Resource-aware inference under latency, memory, and energy constraints
• Human-centered adaptation for smart homes, labs, and connected devices

Motivation

Intelligent IoT systems often need to react to local context quickly and safely. EdgeSLM investigates how compact language models can bring semantic reasoning closer to sensors, devices, and users, making IoT agents more responsive, private, and practical for real-world deployment.