Digitalisation of Biomass Exploration: A Case Study of Biomass Feedstock Classification

Abstract

Biomass is one of the renewable energy (RE) sources with high prospects in the clean energy strata. The diversity of its application and sources have made digitalisation of its exploration presents an array of attractive opportunity to deploy different categories of biomass in a manner that mitigates climate change, advances economies and reduces power dependency. This study discusses the drivers of digitalisation in bioenergy exploration with the associated opportunities and challenges along the bioenergy value chain. As proof of concept, a case study based on the intelligent classification perspective of digitalisation of biomass was discussed. Two classifiers: Sparse Random Error-Correcting Output-based Support Vector Machine (SRECO-SVM) and Euclidean distance-based k-Nearest Neighbour (KNN-EUM) were elaborated while the procedures for the model development were highlighted. Relevant performance indices were applied to evaluate the models developed. Most significantly, the Accuracy, Sensitivity, Specificity were 0.77, 0.81, 0.97 respectively for SRECO-SVM at the computational time (CT) of 20.41 secs while 0.55, 0.56, 0.96 respectively were reported for KNN-EUM at the computational time (CT) of 20.40 secs. Some other composite metrics, which include G-means, F1-score, Mathews correlation coefficient (MCC), Discriminant power (DP) were reported. On the overall, a synergistic relationship is needed between the data analytics, artificial intelligence and other blockchain technologies in order to unleash the full benefits of digitalisation.