Vega: LLM-Driven Intelligent Chatbot Platform for Internet of Things Control and Development

Area of Science:

• Artificial Intelligence
• Internet of Things (IoT)
• Embedded Systems (ESys)

Background:

• Large language models (LLMs) have transformed Natural Language Processing (NLP).
• Their application in Internet of Things (IoT) and embedded systems (ESys) is underexplored.
• Existing IoT interfaces often demand technical expertise, limiting user accessibility.

Purpose of the Study:

• To present Vega, a modular system using LLMs for natural language control of IoT devices.
• To demonstrate intuitive interaction with hardware via LLM interpretation of user commands.
• To reduce the programming complexity for IoT and ESys applications.

Main Methods:

• Developed a system with a physical circuit (sensors, actuators), a Raspberry Pi (RPi) control server, and a web application with LLM logic.
• Users interact via natural language; LLM translates commands for RPi execution.
• RPi executes instructions on the physical circuit; outcomes are relayed to users through LLM-generated responses.

Main Results:

• Vega successfully interprets and executes natural language commands for IoT devices.
• The system handles complex and conditional logic without requiring additional RPi coding.
• Empirical evaluation across various tasks and scenarios validates system performance.

Conclusions:

• LLM-driven IoT control bridges the gap between complex functionality and user-friendliness.
• Vega offers a pathway for more creative and intelligent IoT applications.
• Provides insights into designing and implementing LLM-integrated IoT interfaces.