Digital Twin for Condensation Management during Cold Storage of Fruits

• Applicants: Dr.-Ing. Reiner Jedermann; Dr.-Ing. Pramod Mahajan
• Subject Area: Plant Cultivation, Plant Nutrition, Agricultural Technology ; Measurement Systems ; Plant Physiology
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 521409147

In order to satisfy consumer demand for fresh fruit throughout the year, apples are kept in cold storage under controlled atmosphere for several months. During this period, up to 10% of the fruit is lost, mainly due to moisture and, thus, weight loss but also due to decay, often related to growth of microorganisms under high humidity conditions leading to condensation and wet fruit surface. The mechanisms leading to condensation in cold storage are complex. Models for quality changes in agricultural products as a function of temperature, humidity, and flow conditions in storage rooms and packaging are currently limited to simulation studies. However, it is largely unresolved how to correct the models based on deviations between prediction and actual measured values. Such an update process is essential to make best use of the 'live' sensor data. Accurate management of the condensation conditions and the period during which the surface is wet, therefore, has a huge potential to reduce such losses. The goal of this project is to make the best use of all controllable parameters such as temperature difference at the evaporator, air volume and on/off cycles in combination with sensor data to achieve optimal storage conditions with a beneficial degree of condensation. The concept of digital twins has already been applied successfully in other industrial processes. Nonetheless, applications in agriculture are few and focus mostly on the remote monitoring and less on sensor-based control. The intended development of a digital twin for monitoring and controlling condensation conditions in cold storage facilities starts with development of a detailed predictive model for condensation on fruit and the resulting quality changes. The models will be parameterized and validated by measurements in a climate chamber with full-size apple bins. The models will be transformed into state estimators to allow for continuous correction using live sensor data. After the error propagation analysis, they will be integrated into the digital twin platform. In addition to remote monitoring, the digital twin will in particular allow to predict quantities that are difficult to measure directly, such as condensation at inaccessible locations, duration of wetting and its spatial distribution. Virtual experiments will be used to test the effects of planned interventions such as changes in cooling parameters before they are applied to the real cold store. An automatic assistant will select between different interventions to achieve optimal cooling and condensation conditions. The developed digital twin will serve as a template for further applications in food logistics.

• DFG Programme: Research Grants