Air quality is a vital concern globally, and Sri Lanka, according to WHO statistics, faces challenges in achieving optimal air quality levels. To address this, we introduced an innovative IoT-based Air Pollution Monitoring (APM) Box. This solution incorporates readily available Commercial Off-The-Shelf (COTS) sensors, specifically MQ-7 and MQ-131, for measuring concentrations of Carbon Monoxide (CO) and Ozone (O3) ,Arduino and "ThingSpeak" platform. Yet, those COTS sensors are not factory-calibrated. Therefore, we implemented machine learning algorithms, including linear regression and deep neural network models, to enhance the accuracy of CO and O3 concentration measurements from these non-calibrated sensors. Our findings indicate promising correlations when dealing with MQ-7 and MQ-131 measurements after removing outliers.

Municipal solid waste deposition in metropolitan areas has become a major concern that, if not addressed, can lead to environmental degradation and possibly endanger human health. It is important to adopt a smart waste management system in place to cope with a range of waste materials. This research aims to develop a smart modelling method that could accurately predict and forecast the production of municipal solid waste. An integrated convolution neural network and air-jet system-based framework developed for pre-processing and data integration were developed. The results showed that machine learning algorithms could be used to detect different types of waste with high accuracy. The best performers were obtained from neural network models, which captured 72% of the information variation. The method proposed in this study demonstrates the feasibility of developing tools to assist urban waste through the supply, pre-processing, integration, and modelling of data accessible to the public from a variety of sources.

Assessing the dynamics and patterns of Land Use and Land Cover (LULC) and its transformation is an important practice of urban planners and environmentalists for a variety of applications, including land management, urban climate modeling, and sustainability of any urban region. Monitoring changes in LULC using geospatial techniques can help to identify areas at risk for indefensible land use, low-grade environment, and especially for sustainable urban planning. This study aims to analyze the changing pattern, dynamics, and alteration of LULC using Google Earth Engine (GEE) and Machine Learning Applications for the years 1991, 2001, 2011, and 2022 in the Varanasi City Development Region (VCDR). The LULC classification was divided into seven classes using random forest classification, and Landsat-5(TM) and 9(OLI-2) satellite data were used. Saga GIS has been utilized for the detection of LULC change during the 1991-2022 period. For validation of classification results, accuracy assessment was estimated using error matrices and through user, producer, and overall accuracy estimation. The Kappa statistics were applied for the reliability of the accuracy assessment result. As a result, the built-up area increased by 507.8 percent, and other classes like agricultural, barren, fallow land, and vegetation cover rapidly declined and altered into concrete areas over the period. Water bodies and river sand classes have been slightly converted into different classes. The finding explains that 114.8 km2 of fertile agricultural land, 14.81 km2 barren land, and 12.93 km2 of vegetation cover transformed into impervious surface, which is unsustainable and causes various problems like food scarcity, environmental degradation, and low quality of urban life. This study can be a useful guide for urban planners, academicians, and policymakers by providing a scientific background for sustainable urban planning and management of VCDR and other cities as well.

Biodiesel has been receiving increasing attention because of its fuel properties and compatibility with petroleum-based diesel fuel. Therefore, it is necessary to measure the engine performance and exhaust emissions of engines using petroleum-based diesel fuel and biodiesel blends. The main goal of this study is to investigate the capability of several machine learning (ML) techniques including artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), general regression neural network (GRNN), radial basis function (RBFN), and support vector regression (SVR) for predicting performance and exhaust emissions of the diesel engine fuelled with biodiesel blends. The case application is a Hyundai D4CB 2.5 engine together with B0, B10 and B20 biodiesel blends which are popularly used in Vietnam. The engine process parameters are used as inputs and the outputs include predicted torque and NOx emission. Different predicting models based on ML techniques are developed and validated. The performance of each model is evaluated and compared using root mean squared error (RMSE), mean absolute percentage error (MAPE), mean absolute error (MAE), and correlation coefficient (R). The obtained results indicate that SVR can be used to develop the model for the prediction of performance and exhaust emissions. The study also provides a better understanding of the effects of engine process parameters on performance and exhaust emissions.