Digital Twins in Agriculture: Transforming Farm Management with Virtual Models

Abstract

Digital twin technology, understood as a dynamically updated virtual representation of a physical asset synchronised through continuous bidirectional data exchange, has moved from manufacturing into agricultural production over the past decade. This review examines how digital twins are reshaping farm management across crop production, controlled-environment agriculture, livestock husbandry, machinery maintenance, and agri-food supply chains. It synthesises the conceptual architecture underpinning agricultural digital twins, the sensing and computational technologies that make them possible, and the maturity levels that distinguish a static digital model from a fully synchronised, prescriptive twin. The review traces applications spanning precision irrigation, soil quality assessment, greenhouse climate control, animal health monitoring, and traceability across supply chains, drawing on evidence from cattle, dairy, poultry, and horticultural systems. Persistent barriers are identified, including data interoperability, cybersecurity, the cost of sensor infrastructure, and uneven adoption among smallholder and resource-constrained farms. The review concludes that although digital twins offer measurable gains in resource-use efficiency and decision support, their transition from research prototype to operational farm tool depends on open data standards, affordable connectivity, and closer collaboration between agronomists, engineers, and farmers. Priorities for future research include federated and privacy-preserving twin architectures, tighter integration with artificial intelligence for predictive and prescriptive decision-making, and participatory design that keeps the farmer at the centre of the virtual-physical loop.