GREEN SENSORS AND SMART SERVICES FOR THE OPTIMIZATION OF AGRI-FOOD SUPPLY CHAINS IN A VIEW OF INDUSTRY 4.0: GREATER SUSTAINABILITY OF PRODUCTION, BUSINESS COMPETITIVENESS AND REDUCTION OF FOOD WASTE

The 12th Sustainable Development Goal (SDG) focuses on the concept of sustainable consumption and production, with a particular focus on reducing food loss and waste (Goal 12.3) by 2030. In order to achieve this in the fresh-cut chain, it is necessary to implement systemic improvements, including advanced agronomic practices, optimised harvest timing, effective post-harvest storage and innovative packaging. Green technologies such as optical sensing have the potential to enhance quality, extend shelf life and reduce environmental impact. However, their high costs present a significant barrier to adoption, as multiple devices are often required to address the complexities of supply chains. Research is addressing this challenge by developing compact, and low-cost imaging solutions for both pre- and post-harvest monitoring. This doctoral research project concerns the application of non-destructive imaging techniques to assess the quality of some fresh and fresh-cut products, in terms of freshness and shelf life. Additionally, it involves the development of customized optical devices in a view of Industry 4.0, with the objective of enhancing the sustainability of the supply chain and reducing food loss and food waste. Two distinct measurement configurations were employed at laboratory scale, each with a different application objective. The first configuration concerned a cost-effective IoT hyperspectral prototype, that has been designed and successfully tested. The prototype operated within the spectral range of 400 nm to 1000 nm and has been developed using commercial equipment and a 3D-printer. It has reached satisfactory results for the current stage of technology development, which is estimated to be at a TRL (Technology Readiness Level) of 2. The prototype exhibited the capacity to differentiate between healthy and damaged tissue and to be sensible to variations in leafy vegetables over time, thereby demonstrating the capability to detect changes in the product as it undergoes decay. The prototype was designed for future integration at critical points in the production chain. Instead, the second configuration was proposed with the aim of monitoring the quality level of packaged products, thereby improving their management at the sales point and providing consumers with information on the true quality level of the products. This study demonstrated the potential of multispectral imaging for non-destructive monitoring of fresh-cut lettuce deterioration through packaging. The application of chemometric methods enabled effective identification and classification of fresh samples. However, the classification accuracy for subsequent days was found to be variable due to differences in decay rates between sample groups and light reflections from packaging. In conclusion, this PhD project has demonstrated the benefits and potential of imaging systems as an effective and sophisticated tool for quality assessment in the food industry throughout the production chain, as well as one of the key tools for reducing food loss and waste. However, continued research and development is essential to refine these technologies and make them accessible and adaptable for wider applications. By integrating these tools into sustainable practices, the agri-food sector can make significant progress towards achieving global sustainability goals. Keywords: Green technology, food quality, imaging, industry 4.0, sensor design, cost-effective, chemometrics.