Sensores de Espectroscopía de Infrarrojo Cercano como dispositivos IoT para la lucha contra el fraude en el sector del aceite de oliva virgen

Virgin olive oil represents one of the fundamental pillars of the Mediterranean diet and one of the main economic drivers of the Spanish agri-food sector. Its commercial value, closely linked to its quality and authenticity, is often compromised by fraudulent practices such as adulteration, substitution, or mislabelling. In response to this issue, both the International Olive Council (IOC) and the European Union have highlighted the need to incorporate innovative technologies that strengthen control systems and ensure the integrity of virgin olive oil throughout all stages of the production and marketing chain. Near-Infrared Spectroscopy (NIRS) emerges as a rapid, non-destructive, sustainable, and cost-effective digital analytical alternative for quality control in agri-food products. This technique enables the generation of a spectral fingerprint that is representative of the chemical and physical composition of the sample, thus allowing the simultaneous estimation of multiple parameters with high precision and reproducibility, while significantly reducing analysis time and associated costs compared to traditional methods. Recent advances in the miniaturization, portability, and connectivity of NIRS sensors have extended their use beyond the laboratory, thereby facilitating their integration into production environments, industrial lines, and even mobile devices, opening new possibilities for real-time process monitoring. The main objective of this Doctoral Thesis was to evaluate the potential of NIRS technology for the quality control and authenticity assessment of virgin olive oil. For this purpose, NIRS models were developed for the prediction of physicochemical and sensory parameters associated with virgin olive oil quality, such as acidity, peroxide value, specific extinction coefficients (K232 and K270), waxes, ethyl esters, and polyphenol content. In addition, classification models were built to reliably discriminate between the different commercial categories of the product—extra virgin, virgin, and lampante—thus contributing to product authentication and protection against fraud. The development of these models was carried out using instruments with different scientific capabilities, ranging from laboratory equipment (at-line) to portable devices. The results obtained have demonstrated the feasibility of NIRS technology, and in particular, have highlighted the high capability of portable instruments to provide reliable analytical information. These findings are of great interest for facilitating access to and adoption of this technology among small and medium-sized producers, as well as for its potential use by inspectors and control authorities. As a result of the progress achieved, this research culminated in the development of a cloud-based digital platform integrating NIRS technology with web tools and mobile applications. This platform was designed to automatically manage predictions, traceability, and digital labelling of virgin olive oil based on spectral data obtained by the sensors, representing a decisive step toward the digitalization of quality and authenticity control in the olive oil sector. The developed infrastructure provides an intelligent environment that enhances transparency, efficiency, and evidence-based decision-making throughout the entire virgin olive oil value chain.