Agricultural Intelligence: A Technical Review Within the Perception-Decision-Execution (PDE) Framework

Artificial intelligence (AI) is transforming modern agriculture from experience-driven practices into data-driven, intelligent production paradigms. Within our proposed Perception-Decision-Execution (PDE) framework, this paper reviews AI technology advances from year 2015 to 2025 for agricultural intelligence. At the Perception level, we highlight progress in environment sensing systems, particularly unmanned aerial vehicle (UAV) and multi-modal monitoring platforms, for crop disease/pest detection, growth monitoring, and abiotic stress assessment. At the Decision level, integration of heterogeneous data sources, including meteorological recordings, soil measurements, remote sensing (RS), and market information, enables advanced analytical tasks, such as yield prediction, early pest/disease warning, precision irrigation and fertilization planning, and crop management optimization. And at the Execution level, agricultural robots equipped with simultaneous localization and mapping (SLAM) and deep reinforcement learning (RL) facilitate precision spraying, autonomous harvesting, and unmanned field operations. Collectively, AI technologies have demonstrate substantial potential across the PDE chain of agricultural production, while significant challenges persist, such as heterogeneous data fusion, limited model generalization across diverse environments, complex system integration, and high hardware and deployment costs. Future directions are discussed from the perspectives of lightweight model design, cross-platform standardization, enhanced human-machine collaboration, and deep integration of emerging AI paradigms to support scalable, robust, and autonomous agricultural intelligence systems.