Real-Time Gas Monitoring and Anomaly Detection in Petroleum Industry Using IoT and Machine Learning

Abstract

The petroleum industry faces significant safety challenges due to the presence of toxic gases, which pose serious health risks and potential hazards. To address this issue, an innovative solution with the help of the Internet of Things (IoT) and machine learning has been developed. This solution involves an IoT-driven bot equipped with ESP8266 and Raspberry Pi Pico W microcontrollers, designed to monitor and detect toxic gases in real-time. The bot integrates sensors, including MQ-2, MQ-3, and MQ-135, which detect harmful gases such as methane, propane, alcohol, and ammonia. The ESP8266 provides Wi-Fi capabilities, allowing the bot to connect to the internet and transmit data, while the Raspberry Pi Pico W handles sensor data processing. Controlled via the Blynk IoT application, this setup enables remote operation and real-time monitoring. As the bot navigates through the petroleum facility, it collects gas concentration data, which is sent to a Google Spreadsheet for storage and analysis. This data is processed using machine learning algorithms such as Isolation Forest and One-Class SVM, effective in anomaly detection. These algorithms analyze the data to identify unusual patterns or spikes in gas concentrations, indicating potential leaks or hazardous conditions. Upon detecting anomalies, the system triggers alerts to notify personnel, enabling prompt action to mitigate risks. This approach enhances safety by providing continuous monitoring and demonstrates the potential of IoT and machine learning to revolutionize workplace safety in high-risk environments, significantly improving safety protocols and protecting both workers and the environment.