The Effectiveness of AI: Feedback on the Project to Counteract Food Waste

Background: Project-Based Learning (PBL) has proven effective in developing students’ knowledge and skills. This approach was applied with engineering and Bachelor students to tackle an international challenge focused on resource preservation. The paper presents the technical validation of an AI-based system for automatically sorting potatoes: those deemed “good” are for human consumption, while “bad” ones (with black spots) are redirected for animal feed. The solution emerged from a study of the project’s functional specifications and eight months of development. The initiative received support from industrial and institutional partners. Methods: The study utilized a PBL approach to develop an AI-driven solution for automatic potato sorting. Results: Technical development involved implementing a Convolutional Neural Network (CNN) for image classification. A SICK Inspector vision camera was used for real-time image processing. A dataset of 808 grayscale potato images was created for training. The trained model was embedded in the camera for autonomous classification. Performance metrics such as precision, recall, and accuracy were used to validate the sorting model. A Proof of Concept (PoC) including a range of industrial solutions (Internet of Things (IoT)) was thus validated. Conclusions: This internationally awarded project has demonstrates that AI can be effectively applied to reduce food waste by automating the sorting of agricultural products. This approach opens up opportunities for large-scale industrialization, particularly in the agri-food industry.