Embodied intelligence offers a system-level paradigm for improving the adaptability of robotic harvesting in complex horticultural environments by coupling perception, decision-making, and physical interaction. This review systematically analyzes recent advances in embodied intelligence for robotic harvesting. A search of the Web of Science Core Collection covering January 2020 to August 2026 identified 294 records, of which 100 studies were included after screening. Based on the reviewed literature, a literature-grounded Four-Tier framework is proposed, comprising Multimodal Active Sensing, Semantic Context Recognition and Understanding, Knowledge-Driven Decision-Making and Experience Evolution, and Skilled and Compliant Execution. The review shows that embodied approaches have improved individual capabilities across perception, decision-making, and manipulation; for example, representative studies reported over 12% improvement in detection accuracy, reduction in fruit damage from 8.2% to 2.0%, and 89% retention in continual learning. However, reliable autonomous harvesting remains constrained by cross-module uncertainty, real-time crop–robot interaction, continual adaptation, and the lack of standardized benchmarks. Future research should emphasize physically grounded multimodal closed-loop integration, experience-driven adaptation, standardized evaluation, and crop–robot co-design.
Yue Sun, Wei-Dong Jia, Yun-Fei Wang et al.· Applied Sciences· 0 citations
For the problem that, under the soybean-maize strip intercropping pattern, the installation of anti-drift shields changes the mass and inertia characteristics of the boom of a shielded sprayer, making it prone to severe vibration under complex field excitations and thereby affecting application uniformity and operational safety, a semi-active boom vibration damping system based on a magnetorheological damper was designed and tested. First, an equivalent mechanical model of the boom considering added mass and stiffness variation was established, and the characteristics of excitation sources such as road spectra and start-stop impacts were clarified. Second, based on the Bouc–Wen model, the magnetorheological damper was selected and its output force boundaries were designed. The stroke was determined to be 80–110 mm, and the maximum damping forces in the compression and rebound strokes were 1.15 kN and 3.44 kN, respectively. Furthermore, a semi-active vibration damping system with an LK3U-14MT PLC as the core controller was developed. Finally, simulation analysis, free-vibration decay tests, obstacle-crossing tests, and field operation tests were conducted to verify the vibration damping performance of the developed semi-active system. The field test results showed that, at typical operating speeds of 3–5 km/h, the semi-active vibration damping system reduced the maximum vertical amplitude at the boom tip by 28.6–43.1% and the maximum inclination angle by 12.2–55.6%, effectively improving the attitude stability and roll resistance of the boom of the shielded sprayer.
