Review of machine learning applications for predicting the quality of biomass briquettes for sustainable and low-carbon energy solutions

This review discusses how Machine Learning has been applied to predict the quality of biomass briquettes produced from agricultural and municipal solid organic waste, which are crucial for advancing green and low-carbon energy solutions. Traditional methods of assessment of briquette quality involve destructive laboratory experiments, do not favor sample reuse, are time-consuming, and labor-intensive, posing barriers to efficient production. This paper reviews literature on various Machine Learning models applied for predicting and optimizing briquette quality parameters, including combustion, physical, and emission properties. Several Machine Learning models have shown promising results in predicting and optimizing these key parameters for example, a Random Forest model with R2 of 0.9936 in deformation energy prediction and Artificial Neural Networks with R2 of 0.8936 in the prediction of impact resistance. By enhancing the accuracy and efficiency of briquette quality predictions, Machine Learning algorithms contribute to the development of high-quality biomass briquettes, thereby creating sustainable and low-carbon energy systems. This review points to critical literature gaps regarding model generalizability across diverse biomass feedstocks and integration of broader quality parameters. Addressing these gaps will advance AI-based solutions, promote greener energy practices, and support sustainable development. The findings are intended to aid researchers, industry professionals, and policymakers in advancing the production of high-quality biomass briquettes for cleaner energy and sustainable development.