Abstract

The paper considers a digital twin as an information system that maintains an up-to-date digital representation of a physical agricultural asset or process using sensor data, models, and operational records. A layered architecture is proposed: physical layer, data acquisition and normalization, state model, analytics, and a decision/feedback loop. The distinction between a digital model, a digital shadow, and a fully coupled digital twin is emphasized. Key barriers include incompatible data formats, unstable field data, synchronization complexity, and insufficient representation of agronomic context. Implementation should start with a clearly bounded object such as a greenhouse, irrigation system, machine, or technological operation and then scale gradually.