Nanofibers and Artificial Intelligence: A Synergistic Approach for Next-Generation Sensors

Abstract Emerging materials like cannabidiol (CBD) exhibit photoactive properties, including distinct absorption and emission in the UV spectrum. However, CBD’s practical applications in sensor technology are hindered by its susceptibility to environmental degradation, low thermal stability, and limited solubility in conventional solvents. To address these challenges, we present a novel approach that integrates CBD into a polymer matrix via a single-step electrospinning process. This technique not only stabilizes CBD but also enhances its functionality by creating a high-surface-area composite material with excellent porosity. The resulting CBD-polymer nanofibers harness CBD’s ability to absorb UV light and efficiently transfer energy to the surrounding matrix. To enhance the detection capabilities of the system and minimize signal interference, we employ various machine learning (ML) models, namely k-nearest neighbors, neural networks, and Gaussian process. These models enable reliable differentiation between the presence and absence of UV variation. 5-fold cross-validation is conducted to assess the models’ performance. In particular, accuracy, precision, recall, and F-1 scores are computed, resulting in testing scores of around 65% and higher for all methods. This is promising and paves the way for further analysis to design and incorporate a tailored classifier for the signals collected with this novel UV sensor, which will result in accurate UV light detection.