Monitoring the air quality is very important because of elevated pollution levels causing health hazards. Many research related studies are carried out to estimate the pollutants in the air quality in metropolitan cities and areas near thermal power stations are inconsistent because the measured data for the multiple pollutants are not accurate. Though some mathematical models are available, there exists a research gap because of the lag existing in their capability to detect the multi-pollutant concentrations. Hence, a smart drone can be constructed to detect and measure 7 poisonous gases causing air pollution in various metropolitan cities and regions near thermal power plants. The quality of air can be monitored using drones in surrounding areas, near thermal power plants like Neyveli Lignite Corporation (NLC), Neyveli, Tamil Nadu, for solid waste estimation in the dump yards and also emission of various flue gases like Carbon Monoxide (CO), Carbon dioxide (CO2), Nitrogen Oxides (NOx) and Sulphur Dioxide (SO2). Currently, simulation studies reveal that four gas emissions related to air pollution are forecasted and estimated on using real time data from the sensors on a small scale. This data was used for analysing the training curve of the Radial Basis Function Network (RBFN). The proposed RBFN model was capable of monitoring and forecasting the intensities of various pollutants present in the atmosphere sample used for pollution analysis. Once the training and testing is over, the estimated and forecasted values of the pollutants to measure the air quality in various places can be used to avoid pollution or minimize the pollution levels by ensuring complete combustion. The prototype model of RBFN yielded an RMS value of 0.0001 for measuring the four types of air pollutants present in the flue gases from NLC (industrial area) for estimation and forecasting which will support the urban sustainability leasing to planning and development of pollution-free smart cities.