Digital twin-based approaches for agricultural machinery damage prediction and maintenance: A review

The reliability of agricultural machinery is increasingly constrained by harsh operating environments, complex dynamic loads, and evolving failure mechanisms, posing critical challenges to agricultural production efficiency and system resilience. Traditional maintenance methods, often reactive and resource-intensive, are insufficient to meet the demands of modern precision agriculture. There is a limited comprehensive review of how digital twin-based approaches can overcome these challenges by integrating data-driven models, intelligent prediction algorithms, and real-time maintenance decision-making strategies. Therefore, this paper reviews digital twin-based strategies for agricultural machinery damage prediction and maintenance optimization. Three key elements are analyzed: (1) Numerical modeling approaches for simulating mechanical behavior and predicting damage evolution under diverse operational conditions; (2) Advanced fault diagnosis techniques integrating machine learning algorithms and multi-source sensing to enable real-time monitoring, condition assessment, and early anomaly detection; (3) Additive manufacturing (AM) technologies for the rapid repair and reinforcement of damaged components, supporting efficient lifecycle management. By integrating numerical simulation, intelligent diagnostics, and additive repair into digital twin frameworks, a predictive, closed-loop maintenance paradigm is established, enabling proactive interventions and enhanced operational continuity. Key challenges, including material and process limitations, portability and equipment adaptation, as well as model fidelity and real-time integration, are discussed. This review aims to provide a systematic reference for advancing digital twin technologies in agricultural machinery, which outlines future directions toward intelligent, sustainable, and resilient agricultural systems.