Preliminary Phases of Implementing a Digital Twin Solution in the Food Industry: a Case Study

In the current era of digital transformation, food companies are increasingly tasked with developing systems that enhance sustainability and operational efficiency. Small-to medium-sized food plants, in particular, often rely heavily on experience-based procedures, which can be financially suboptimal and lack the flexibility needed to adapt to changing demands. The digitalization and automation of production processes pose a critical challenge, offering significant opportunities to improve quality, competitiveness, and sustainability across the sector. Among the emerging technologies, the digital twin stands out as one of the most promising solutions. This paper presents results from a broader project aimed at enhancing productivity and efficiency while aligning with quality and sustainability objectives through the implementation of digital twins. Focusing on the industrial production process of vegetable broth, this study examines the initial phases of implementing a digital twin. These phases involve adopting a preliminary protocol that enables the reconstruction of material and energy flows within the process, the development of a Process Flow Diagram (PFD), the collection and analysis of process variables, and the resolution of a system of energy and material balance equations. This methodology addressed information gaps caused by limited sensor availability, facilitating the creation of a comprehensive dataset essential for digital twin implementation. Building on this dataset, the subsequent phase applies statistical methods, such as data reconciliation, to minimize errors and further enhance data accuracy. The refined dataset is then integrated into specialized simulation software, enabling the implementation of the digital model and the identification of optimization solutions. Additionally, the study highlights that integrating advanced sensor systems directly within the plant yields higher-quality data compared to traditional technical plant data or measurements obtained from offline sensors. This underscores the importance of investing in modern sensor technology to support the successful adoption of digital twin solutions in the food industry.