A preliminary implementation of a retrofitted AI demand forecasting system for Digital Twin-based production scheduling in the food sector

The food industry is increasingly focused on enhancing demand forecasting tools to align production with consumer needs better while minimizing waste. This study investigates the integration of Artificial Intelligence (AI) and Digital Twin (DT) technologies to improve decision-making flexibility in the food sector. Specifically, a demand forecasting tool utilising a continuous XGBoost algorithm was developed, serving as the input for a Digital Twin system designed for a real Italian company in the fresh fruit production scheduling sector. The forecasting system operates in a feedback loop where the forecasted demand is compared with actual orders, enabling real-time production re-scheduling. The results highlight the potential of AI and DT technologies, combined with retrofitting and adaptive correction, to drive continuous optimization of the decision-making process for sustainability.