Indoor pollution has become a concern all over the world. The means of source of fire in low-income and middle-income countries use firewood, charcoal, and liquefied petroleum gas (LPG) for cooking daily. According to the world health organization, an estimated 1.6 million people die annually due to indoor pollution. The Institute for Health Metrics and Evaluation (IHME) found out that 6% of deaths in low-income countries are caused by indoor air pollution. Indoor air pollution increases the risk of contracting noncommunicable diseases such as heart disease, stroke, lung cancer and chronic obstructive pulmonary disease (COPD), lung cancer, and pneumonia among others.

Among the main sources of household air pollution are harmful gasses from wood fuel and Liquefied Petroleum Gas (LPG). Even though, attempts have been made to come up with solutions using different technologies to monitor the pollution levels and give appropriate alerts. There is a need to develop a solution that will not only monitor but also predict pollution levels so that corrective measures may be taken early enough. This study aims at developing a prototype for a household air quality monitoring system and pollution prediction system using Artificial Intelligence (AI).

The proposed solution will involve the installation of the system in the house with the use of a low-cost Arduino-based microcontroller that incorporates sensors to monitor carbon monoxide (CO) and particulate matter and environmental sensors to monitor temperature and humidity. The collected data will be sent to Cloud via Global System for Mobile (GSM) for data storage, processing, and analysis. The alerting system will be developed along with the prediction system using Artificial Intelligence. Pollution early warning alerts are generated appropriately. The expected results from the study are to collect data from a kitchen, alert people using the kitchen when the air quality is poor in real-time, and also the prediction of air quality so as to minimize and later avoid the exposure of poor air quality in households.

This work deployed sensors in the household connect them with Arduino UNO and send the collected data to a cloud platform known as Thingspeak with the use of a GSM module, the data was further used on air quality prediction using Machine Learning algorithms, and air quality widget will clearly show the level of air pollution in colors.

This system will help reduce the health risks relating to indoor air pollution and also provide data to be used by the government, NGO environmental bodies, and agencies.