An IoT-Enabled Real-Time System for Detection and Analysis of Toluene using Machine Learning

Abstract

Background / Objective: According to World Health Organization (WHO) reports, Toluene is one of the volatile organic compounds (VOCs) that is released by industries and vehicles causing serious health concerns, including brain damage and respiratory problems. To reduce health risks, maintain regulatory compliance and help the regulatory agencies to frame the future policies, toluene concentrations must be accurately and consistently monitored. However, traditional detection methods, including gas chromatography-mass spectrometry (GC-MS), lack real-time monitoring capabilities and are very expensive from installation and maintenance point of view. Methodology: The presented work suggests a portable, low-cost, real-time toluene detection system that integrates inexpensive sensors with IoT architecture for real-time data Collection and monitoring on the Amazon Web Service (AWS) cloud platform. The system deals with Machine learning (ML) based analysis for trend identification and interpretation. Findings: The strong correlation between VOC grade and Toluene concentration (0.98) shows that the developed device is highly capable of detecting the toluene concentration. The dataset can be useful to the external agencies for the pollution audit in industrial settings. Novelty / Improvement: The device is developed using the low cost and portable sensors, collects the real time data using the AWS cloud. Therefore, this device can be easily installed in industrial settings that can ensure increased workplace safety and regulatory compliance by providing an affordable and scalable environmental monitoring solution.