Akademik Veri Yönetim Sistemi
2025 Innovations in Intelligent Systems and Applications Conference (ASYU), Bursa, Türkiye, 10 Eylül 2025, (Tam Metin Bildiri)
This paper describes the design and implementation of an ESP32-S3 based IoT device that integrates real-time gas safety monitoring with an AI-powered voice assistant, featuring enhanced user feedback and conversational context. The system’s primary aim is to combine hands-free voice interaction with a Large Language Model (LLM) and real-life MQ-2 gas detection for environmental safety monitoring. The prototype demonstrated robust performance in handling voice queries with contextual memory and environmental monitoring. This work illustrates a cost-effective and adaptable approach to developing multifunctional solutions by combining ESP 32 with modern cloud-based AI services and comprehensive user feedback mechanisms, suitable for applications in home automation, safety, and accessibility. Simultaneously, the system conducts environmental monitoring using an MQ-2 gas sensor. Upon detection of gas levels exceeding a predefined threshold (e.g., an ADC reading of 800), the system triggers a local voice alert through its speaker, activates a buzzer, and pushes notifications through both Telegram and the Blynk mobile dashboard, ensuring timely local and remote awareness.Index Terms—ESP32, IoT, Voice Assistant, Gas Detection, LLM, Conversational AI