AI-Based IoT Gas Leakage and Fire Detection
—Gas leakage, fire incidents, and abnormal envi- ronmental conditions create safety risks in homes, laborato- ries, workshops, and industrial settings. This paper presents RiskSense, an ESP32-based Internet of Things (IoT) prototype designed to monitor gas-sensitive readings, flame status, tempera- ture, and humidity. The prototype combines an MQ-2 gas sensor, a digital flame sensor, and a DHT11 sensor with local indicators, a buzzer, a motor driver, and the Blynk IoT platform. Sensor readings are processed using configurable thresholdbased logic for local alarms, while status values and alerts can be viewed remotely through a Wi-Fi-connected dashboard. An artificial intelligence (AI) layer based on a Random Forest classifier is implemented to classify the monitored conditions into NORMAL, WARNING, and CRITICAL risk categories. Isolation Forest for unusual sensor-pattern detection is proposed as future work. The paper describes the system architecture, hardware and software design, control logic, the Random Forest algorithm used, implementation methodology, validation plan, limitations, and future work. Since measured experimental results and numerical model-evaluation metrics have not been provided, no numerical performance claims are made. Index Terms—Internet of Things, ESP32, MQ-2, flame detec- tion, DHT11, gas leakage, fire detection, environmental monitor- ing, Blynk, Random Forest, machine learning.
Authors
- Shuyang Zhang (ORCID: https://orcid.org/0009-0003-6605-7245)
Institutions
- Ajeenkya DY Patil University (IN)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-19
- DOI
- https://doi.org/10.5281/zenodo.24371738
- Primary Topic
- Fire Detection and Safety Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00