Vertical farming has emerged as a promising solution to cope with increasing food demand, urbanization pressure, and limited resources and to ensure sustainable year-round urban agriculture. The aim of this review was to investigate the evolving technological landscape and engineering considerations, with a focus on innovative developments and future prospects. This paper presents technological trends in vertical farming, covering advances in sensing technologies, monitoring and control systems, and unmanned systems. It also highlights the growing role of artificial intelligence (AI) in contributing to data-driven decision-making and the optimization of vertical farms. A global perspective on vertical farming is presented, including the current status and advanced technological trends across regions like Asia, the USA, and Europe. Innovative concepts and upcoming enterprises that could shape the future of vertical agriculture are explored. Additionally, the challenges and future prospects of vertical farming are also addressed, focusing on crop production limitations, environmental sustainability, economic feasibility, and contributions to global food security. This review provides guidance on the state of vertical farming, technological advancements, global trends, challenges, and prospects, offering insights into the roles of researchers, practitioners, and policymakers in advancing sustainable vertical agriculture and food security.

Cabbage (Brassica rapa L. ssp. Pekinensis) is an important vegetable crop in the Republic of Korea, due to its essential role in kimchi production. However, labor shortages and an aging population necessitate mechanization to sustain productivity. This study aimed to evaluate the field performance of a cabbage transplanter under development with a commercial transplanter and a cabbage collector under greenhouse conditions. This study evaluated transplanting efficiency, planting performance, and yield of cabbage using seedlings at three distinct age groups (30, 35, and 43 days). A cabbage transplanter (Transplanter A) under development, a commercial model (Transplanter B), and manual transplanting were used for comparative analysis. At harvest, a tractor-mounted cabbage collector was used to collect and pack all the cabbages. Transplanter A demonstrated a forward speed of 1.27 km/h and an average planting rate of 2365 seedlings/h, significantly higher than manual transplanting (513 seedlings/h). The effective field capacity (EFC) ranged from 0.11 to 0.13 ha/h, compared to 0.019–0.028 ha/h for manual planting. While Transplanter A showed a higher missing transplant rate (18.17%) than Transplanter B (7.67%), it maintained consistently lower bad planting rates (2.5–4.5%) compared to Transplanter B (3.3–8.8%). In addition, it produced significantly higher cabbage weights (6070 g/plant) and better root metrics than manual transplanting. The cabbage collector achieved 100% efficiency with no crop damage or contamination. The transplanter under development proved effective for greenhouse use, offering faster operation, better planting accuracy, and higher yields, supporting broader mechanization in Korean agriculture.

Root and tuber vegetables, including potatoes, radishes, sweet potatoes, and carrots, are important components of Korean upland agriculture, yet harvesting and yield determination remain less mechanized and digitally integrated than cereal production. This review integrated contact- and weight-based yield measurement technologies applicable to Korean root and tuber crop harvesting, with emphasis on load-cell systems, conveyor weighing, hopper/bin weighing, impact-force sensing, and integrated multi-sensor approaches. Korean and international studies were checked in relation to sensing principles, harvester integration, calibration, dynamic measurement, spatial yield mapping, and field applicability. The literature shows that direct load-cell-based measurement provides the most physically interpretable basis for harvested mass estimation, while conveyor systems are particularly suitable for continuously conveyed crops such as potatoes and radishes. Hopper and cart weighing provide practical alternatives for semi-mechanized operations, including sweet potato harvesting, whereas impact sensing offers compact measurement but is more sensitive to crop trajectory, vibration, inclination, and mechanical damage. Korean studies have demonstrated promising sensor-level performance for potatoes and radishes; however, field-scale implementation remains constrained by vibration, slope, soil and debris contamination, tare drift, crop transport delay, limited standardization, and differences in harvester architecture. Integration of direct mass sensing with GNSS, IMU, belt-speed sensing, machine-state information, and machine vision provides a promising pathway toward reliable yield mapping. The synthesis indicates crop-specific pathways: continuous conveyor weighing after soil separation for potatoes; dynamically compensated event/impact sensing for radishes; low-impact hopper, cart, or batch weighing for sweet potatoes; and post-cleaning weighing for carrots. The main remaining gap is integrated field validation under realistic throughput, vibration, slope, headland transitions, and field-scale conditions. Future studies should report accuracy together with throughput, calibration burden, spatial-lag error, contamination effects, cost, and field repeatability.

Precision seed metering has evolved from mechanically optimized devices to sensor-based, digitally connected systems. However, research remains fragmented across metering mechanisms, sensing, data processing, communication, and control. This review examined monitoring and closed-loop control throughout the seed-delivery process, from pickup and release to tube passage, furrow placement, and crop emergence. This structured review synthesized 49 primary studies identified through Scopus and the Web of Science Core Collection, with supplementary searching using Google Scholar and literature coverage through 1 September 2026. Optical and photoelectric sensors provide the most direct evidence for seed-passage detection, whereas encoders, pressure, vacuum, motor-current, and vibration measurements primarily indicate machine or actuator conditions. Machine vision provides spatial information and can support independent reference measurements, but requires suitable imaging conditions and computational resources. Capacitive, acoustic, impact-based, and hybrid approaches remain less extensively validated. Reported performance cannot be directly compared across studies due to differences in seed characteristics, mechanisms, operating speeds, event definitions, disturbances, and reference methods. Most studies demonstrate monitoring, alarms, or open-loop optimization rather than complete sensor-to-actuator correction. An engineering framework is therefore proposed to connect measurement targets, evidence maturity, latency, actuator authority, and independent verification. Future research should prioritize standardized testing, disturbance robustness, synchronized multi-row field validation, and measured end-to-end delay. Transparent reporting of correction accuracy, stability, cost, power demand, and interoperability is also required to support deployment and defensible assessment of closed-loop seeding systems.