Swarm-Controlled Agricultural Robots: A Comprehensive Review of Architecture, Stability Strategies, and Field Implementation in Citrus Orchards

The rapid expansion of agricultural robotics is increasingly constrained by complex field environments, operational variability, and the growing demand for coordinated multi-robot tasks, rendering traditional single-robot systems insufficient for modern agricultural production. Although swarm control has emerged as a promising solution, existing studies lack a systematic framework that integrates coordination architecture, agricultural scenario adaptation, and field-level validation. This paper analyzes and highlights the importance of Swarm-Controlled Agricultural Robots (SCAR), and summarizes current research hotspots regarding the integration of SCAR with agriculture. Secondly, we present a structured framework for Swarm-Controlled Agricultural Robots based on three core dimensions: swarm coordination architecture, crop-oriented task adaptation, and environmental robustness design. The proposed model integrates centralized–distributed hybrid control, task-specific parameter optimization, and anti-interference mechanisms tailored to complex orchard environments. Based on this framework, representative application paradigms in precision planting, collaborative harvesting, and intelligent management are analyzed, and stability enhancement strategies are proposed from hardware, algorithmic, and communication perspectives. Long-term field validation conducted in citrus orchards from 2021 to 2025 demonstrates that the proposed framework improves operational efficiency by 45%, achieves positioning accuracy within one centimeter, and significantly enhances environmental adaptability compared with conventional single-robot systems. This research provides a practical foundation for large-scale deployment of agricultural multi-agent systems.