Answers are grounded in this paper and cite their sources.
Forage production currently faces multifaceted pressures, including resource constraints, climate fluctuations, and escalating labour costs. Accurately and promptly acquiring data on forage growth status and translating it into operational prescriptions is crucial for enhancing forage utilisation efficiency and production sustainability. This paper presents a comprehensive review of intelligent forage production, focusing on multi-source perception, biomass and yield estimation, precision operations, and system synergy. To ensure a rigorous synthesis, a structured literature search was conducted across Web of Science, Scopus, and CNKI, covering publications up to August 2026. Firstly, the roles of ground-based, near-surface, unmanned aerial vehicles (UAVs) and satellite platforms in monitoring plant height, canopy structure, spectral characteristics, moisture, and nutrients are compared; this is then followed by an analysis of data acquisition and calibration methods across varying spatial scales and temporal resolutions. Secondly, strategies for biomass and quality estimation—based on vegetation indices, red-edge and short-wave infrared (SWIR) spectroscopy, canopy height and three-dimensional (3D) structure, as well as machine learning—are synthesised. Particular emphasis is placed on the impacts of soil background, spectral saturation, mixed-sward heterogeneity, post-mowing regeneration, and cross-regional generalisation on model accuracy. Furthermore, precision operation technologies, such as variable-rate irrigation, fertilisation, optimal harvest window identification, cutting control, baling, and autonomous navigation, are summarised. The prerequisites for synergy among perception, estimation, decision-making, execution, and operational feedback are also analysed. A key synthesized finding is that structural-spectral feature fusion only yields actionable precision gains when temporal synchronization and spatial scale matching are strictly enforced, and continuous simplified local calibration remains indispensable for cross-regional model generalization. Finally, the integration of open-access multi-source data, the fusion of physical models with artificial intelligence (AI), rigorous external validation, and farmer-oriented low-cost deployment are proposed as key priorities for future research.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: