Novel computational and forecasting strategy for environment quality monitoring using deep learning

Abstract

Air pollution is the serious issue that one must think about and is caused by harmful gases present in the atmosphere such as Carbon Dioxide, Carbon Monoxide, Sulphur Dioxide etc. since level of pollution varies from one place to another. According to WHO (World Health Organization) air pollution is the fifth major cause for deaths after heart diseases, high blood pressure, poor nutrition and tobacco smoking. Monitoring and detection of the amount of harmful gases over particular area can reduce the chances of endanger to human beings and warn to take precautionary measures and do necessary remedies to regulate the emission of poisonous atmospheric gases. The present paper deals with the monitoring of the disastrous gases using gas sensor which is embedded with NodeMCU. The observed levels sent through internet to cloud platforms using MQTT protocols. The data is stored in the ThingSpeak cloud which can be further analyzed from anywhere in world. Data is processed using Machine learning (ML) algorithm called Long Short-Term Memory Network (LSTM) which is the state-of-the-art technique in the field of data analytics and majorly used for data forecasting.