[Special Founding Interview] Asking Kim Dae-hyun, President of the National Institute of Horticultural and Herbal Science, About the Future of Agriculture

South Korea’s horticulture and specialty crop industry has entered a major turning point. Poor crop performance caused by climate change, severe labor shortages due to the aging rural population, and soaring labor and material costs are key factors threatening the survival of farms in the field. Amid this crisis, the National Institute of Horticultural and Herbal Science, which inherited the spirit of Dr. Woo Jang-chun and laid the foundation for horticultural research in Korea, is seeking a breakthrough by putting forward cutting-edge smart agriculture based on artificial intelligence (AI), robots, and data. To mark its founding, this newspaper met with Kim Dae-hyun, president of the National Institute of Horticultural and Herbal Science, which oversees research and development (R&D) in Korea’s horticulture and specialty crop sectors. Together with President Kim Dae-hyun, we discussed the future roadmap for Korean agriculture facing abnormal weather and labor shortages, as well as field-centered ag-tech solutions.

Interviewer: Lee Hyun-woo, Editor-in-Chief
Date: June 15, 2026
Location: National Institute of Horticultural and Herbal Science 

-The National Institute of Horticultural and Herbal Science, which inherited the spirit of Dr. Woo Jang-chun and established the foundation in fields such as horticulture in Korea, marks its 73rd anniversary this year. Please share your thoughts.
“Over the past 73 years, we have steadily contributed to the development of Korean agriculture and industry through variety development, cultivation technology innovation, and solving field problems. At the heart of this lies the spirit and dedication of Dr. Woo Jang-chun, the starting point of our institute and its first director. In the 1950s, Dr. Woo awakened the nation to the importance of seed and horticultural research and laid the research foundation in areas essential to the public’s diet and agricultural sites, including cabbage, radish, flowers, and citrus. As the head of an institution that has inherited this history and tradition, I feel great pride along with a heavy sense of responsibility.
The horticulture and specialty crop sector now accounts for about 26 trillion won of Korea’s agricultural production value, and its stature continues to grow. Even compared with data from five or 10 years ago, it has been on a constant upward path. In the past, the goal was self-sufficiency, but now ‘well-being and health’ have become key keywords, extending beyond fruit and floriculture to healing agriculture and health functional foods using ginseng and specialty crops. As both farmers and consumers are paying attention to the value of health, I am convinced this industry will grow even more explosively in the future. We will respond more proactively to new conditions such as climate change, digital transformation, and changing consumption patterns. On the foundation built by our senior researchers, we will strengthen future-oriented research and create outcomes that provide real help to agricultural sites and industry." 

-What is the core task you most want to focus on during your term?
“For more than 70 years, the institute has led agricultural modernization by developing numerous horticulture and specialty crop technologies. However, if excellent R&D achievements are not absorbed in the field and connected to consumers, they are not living technologies. Therefore, during my term, I intend to devote all efforts to narrowing the speed gap between laboratories and farm sites and to boosting the ‘field penetration rate’ of technology and the ‘creation of tangible results.’ Another focus is future response in preparation for abnormal weather and global warming. I will place emphasis on ensuring that technologies farmers can actually feel in the field—such as smart farms, physical AI, and the development and dissemination of disaster-resilient varieties—are practically distributed.” 

-This year, the Rural Development Administration put forward a work plan under the banner of ‘Agriculture growing bigger through AI convergence, a happier rural community together.’ What are the key cutting-edge AI smart agriculture policies and priority projects the institute is focusing on?
“The core is to use AI, data, and automation technologies to make the entire cycle of horticulture and specialty crops—from production to post-harvest management, pest and disease control, and variety development—more precise and efficient. First, we are focusing on advancing smart agriculture across both open fields and facilities. We are building smart management systems for major fruit trees such as apples, pears, citrus, and grapes, as well as field crops such as cabbage, garlic, and onions. Representative examples include mechanized and automated production technologies for apple orchards, image diagnosis and AI monitoring technologies, low-temperature damage prediction models, and summer cabbage crop observation and forecasting technologies.
We are also spreading data-based precision cultivation technologies in facility horticulture and medicinal/special-use crops. We are developing standard environmental control attribute information for greenhouse horticultural crops and promoting the advancement and dissemination of precision nutrient and water management systems and the Ara Greenhouse, a next-generation greenhouse platform. In the specialty crop field as well, we are developing AI-based smart management technologies such as oyster mushroom growth prediction and ginseng growth modeling.
The digitalization of post-harvest management and distribution is also under way. We are building a smart APC demonstration model, developing quality management technologies linked to production environment data, and conducting research on citrus quality tracking and smart distribution technologies. Fourth is the development of AI-based pest and disease forecasting and precision control technologies. For major items such as chili peppers, peaches, ginseng, and oyster mushrooms, we are focusing on reducing farm damage by advancing early diagnosis, outbreak prediction, and customized control technologies using image analysis and deep learning.
Finally, in the field of variety development, we are pushing in earnest for a transition to digital breeding. We are working to shorten the time required to develop new varieties through image-based phenotypic analysis and integrated analysis of genomic and omics data. In addition, in the field of healing agriculture, we are promoting convergence research such as AI-based emotional analysis, cognitive activation response analysis, and the establishment of a companion plant interaction platform.”

-The most closely watched project in the fruit sector is the ‘establishment of a future-type apple cultivation system.’ What was the background behind launching this project, and what is the ultimate cultivation goal it aims for?
“Apple cultivation in Korea has long developed around the ‘slender spindle’ tree form. However, because the canopy structure is three-dimensional, there have been limits to the mechanization and automation of key tasks such as pruning, thinning, and harvesting. This became a constraint on reducing production costs amid worsening rural aging and labor shortages, and frequent weather disasters such as abnormally high temperatures also made stable production difficult.
Accordingly, we aim to shift the cultivation system based on a ‘planar tree form,’ in which trees are arranged in a flat structure, and to develop mechanized technologies suitable for it, including pruning, pest control, and harvesting. The goal of the future-type apple cultivation system is to produce stable yields and high-quality fruit on the same area with less labor and lower cost. We will dramatically reduce labor, raise productivity and fruit quality at the same time, and present a new standard model for a sustainable apple industry." 

-Structurally, how does the future-type apple cultivation system differ from the existing traditional apple cultivation system? In particular, we are curious about changes in productivity such as reduced labor hours and lower operating costs.
”The future-type apple cultivation system is a production system designed on the premise of mechanization, automation, and smartization. Existing orchards are based on the slender spindle tree form, with large trees and high working heights, so most tasks have had to rely on manpower. By contrast, the future-type system lowers tree size and simplifies orchard structure based on the planar tree form, enabling efficient use of machinery and smart equipment.
The change farmers will feel most strongly is labor reduction. In planar orchards, work routes are simple and working heights are low, so the efficiency of key tasks such as pruning, thinning, and harvesting is high. Reduced labor input leads to lower production costs. In the existing cultivation system, labor costs account for a large share of operating expenses, but the future-type system reduces working time and lowers dependence on manpower, easing the burden of management costs. Initial facility investment costs may be somewhat higher, but in the long term economic feasibility improves. The planar tree form enhances photosynthetic efficiency and improves fruit coloration and quality, making it advantageous for securing stable yields and improving farm profitability." 

-There are growing calls that, for agricultural robots to be used in the field, a digital environment in orchards must first be established. How are you preparing ‘standardization of the orchard digital environment’ and plans to disseminate smart technologies?
“Standardization work for spatial information, growth and environmental data collection systems, and linkage with smart orchard platforms must come first. First, to introduce autonomous driving and farm-work robots in orchards, we are standardizing 3D electronic maps of orchards based on LiDAR and RGB and developing digital twin-based mapping technologies.
In addition, we are building tree-level data by collecting and linking IoT sensors, weather observation networks, and image data in real time, and we plan to standardize these by item and use them as input data for AI-based growth diagnosis and pest-control decision-making. The constructed data will be linked to smart orchard platforms and used to provide orchard-tailored farm-work AI solutions.
The dissemination of smart technologies will shift away from an individual equipment-centered approach to the spread of integrated services based on data and AI. We will build smart agriculture testbeds for apples, pears, grapes, and citrus to develop and demonstrate core technologies such as cultivation environment monitoring, growth diagnosis technologies using image data, and automatic irrigation control that analyzes soil moisture distribution. At the same time, we are collaborating on the development of farm-work automation technologies such as a weeding robot from the National Academy of Agricultural Science and a precision pruning robot from Chonnam National University. Ultimately, we are researching the establishment of an integrated smart farming management platform that manages all these data in an integrated way so that farmers can use farming decision-making services on a single platform.” 

-Mechanization of field crops such as garlic and onions is also an urgent task. How far has Korea come in the mechanization stages for garlic and onions at present?
“As of 2023, the overall mechanization rate for garlic and onions is 67.5%, but the sowing/transplanting and harvesting stages, which actually require the most labor, still remain at low levels. The mechanization rate for garlic sowing is 17.6% and harvesting 59.7%, while for onions the mechanization rate for sowing/transplanting is only 22.7% and harvesting 31.4%.
To overcome this, we are promoting the development of full-process mechanization technologies in cooperation with the National Academy of Agricultural Science, and recently held performance-sharing meetings in Changnyeong (garlic) and Hamyang (onions) to share the technologies. A particularly noteworthy area is ‘non-mulching (uncovered) cultivation technology.’ In conventional plastic mulch cultivation, removing the plastic before mechanical harvesting alone required about five hours of labor and 100,000 won in cost per 10a, reducing mechanization efficiency. To prevent the productivity decline that can occur in non-mulching cultivation, we are advancing cultivation technologies that combine optimized planting density with 5–12% denser planting than conventional practice, labor-saving fertilization technologies such as deep fertilization and ridge fertilization using tractor-mounted fertilizer applicators, and the setting of proper irrigation intervals based on moisture sensors. As for the weed problem caused by the absence of mulch, we are promoting expanded herbicide use for spring weed management by establishing registration test standards for herbicide treatment methods during the spring growing season.”

-What are your future plans for transitioning to AI- and data-based smart agriculture in open fields?
“To transition to smart agriculture in open fields, we plan to systematically collect growth and yield data—such as plant height, number of leaves, dry matter weight, and bulb weight—as well as drone image data in major production areas such as Muan and Hampyeong in South Jeolla Province, and secure a reference database. Based on correlation analysis between multispectral drone images and field data, we will develop garlic and onion growth estimation models and use them as scientific grounds for early response to abnormal weather and pests and diseases, as well as for establishing supply-demand stabilization policies.”

-The biggest obstacle to the spread of smart farms is the burden of initial capital. How much can the ‘Ara Greenhouse Platform’ under development actually lower installation costs for farms? We are also curious about customized alternatives for very small and small-to-medium farms.
“Until now, smart farm policies have been biased toward large-scale advanced greenhouses, creating a high barrier to entry for the 74% of small and medium-sized farms under 0.5 hectares. A representative achievement developed to lower costs is the Ara Greenhouse Platform. This system is a kind of app store that provides interoperability among electrical and communications infrastructure standards and equipment such as sensors and actuators through Internet of Things technology. Once commercialized, it can solve cost problems and make it easy to install AI programs. Existing domestically made integrated environmental control devices cost 25 million won, and imported ones can cost up to 100 million won, but the Ara Greenhouse Platform has dramatically lowered the cost to 14 million won while improving maintenance and convenience.
In addition, we are setting crop-specific input technologies such as ventilation windows, circulation fans, ventilation fans, shading technology, heaters, and multi-layer thermal curtains, while on the software side we support NCPMS for pest and disease forecasting and diagnosis, a weather disaster warning system, and ‘Isagi.’ Starting this year with strawberries and tomatoes, we plan to conduct field demonstrations of the economic feasibility of small and medium-sized smart farm models for a total of 10 crops. If these customized solutions spread in the field, we expect they will produce tangible management improvement effects, reducing operating costs for small and medium-sized farms by 25%, increasing productivity by 30%, and improving income by 20%.”

-Please introduce the current status of the citrus smart farm pilot project now under way and the achievements in digital control.
“For open-field citrus, the proportion of elderly farmers in their 60s or older among growers surged from 33.4% in 2015 to 50.7% in 2024, and the share of women farmers also reached 44.1%, making smartization urgently needed. However, the trees are old, with many over 40 years of age, and the high proportion of dense planting has limited high-quality production and mechanization.
When I took office as head of the Citrus Research Institute in 2022, I keenly felt the need, secured the budget, and laid the groundwork. Open-field citrus also urgently needs tree-form development and unmanned pest control, so we are leveling the ground on a 2,310㎡ site and building a testbed that introduces central leader, two-axis, and four-axis tree forms. We are promoting automatic weather observation, Tyvek soil-cover cultivation, automatic drip irrigation using soil moisture sensors, unmanned monitoring, and unmanned pest control. In the future, we will establish an integrated smart farming management system for open-field citrus. Starting in July this year, we will jointly conduct with industry a demonstration study in citrus orchards of an ‘AI-based smart pest-control robot system’ to introduce autonomous-driving speed sprayers, and in late October we plan to hold an on-site evaluation meeting for the open-field citrus testbed to publicize its potential.”

-If digital transformation is applied to the post-harvest management field, what kind of innovation will come to the marketability of our agricultural products and precision supply-demand management? What is the roadmap for this?
“We are pursuing two strategic directions for the digital transformation of post-harvest management. First is data-based precise quality prediction and smart supply-demand management. Until now, storage and distribution management has focused on meeting quantitative conditions such as temperature and humidity, but going forward we must predict quality changes based on data and adjust shipment timing and volume. We are systematically accumulating data on quality, storage environments, and distribution changes for major horticultural crops, and we are advancing technologies that use AI to predict the point of quality deterioration during storage, the possibility of spoilage, and the marketable distribution period. In production areas, this can help determine the optimal shipping time, and in distribution sites it can reduce losses.
The second is a paradigm shift in the criteria for grading agricultural products. In the past, the focus was on quantitative figures such as size, weight, and sugar content, but in the future we must also reflect qualitative quality factors such as taste, texture, freshness, and consumer preference. To this end, we plan to combine non-destructive quality diagnosis, image and sensor data, and consumer evaluation information to create standards that can select and distribute ‘agricultural products that are satisfying to eat.’”

-Please share a word of encouragement and advice for the readers of Korea Agricultural Technology Newspaper, which launches on July 1, and for Korea’s future tech farmers.
"Agriculture is now rapidly evolving into a new future industry, with advanced technologies such as AI, robots, and digital twins being combined with the foundation of a traditional industry. At this important turning point, the launch of a specialized media outlet connecting agricultural technology and the field is highly meaningful. I hope Korea Agricultural Technology Newspaper will become a valuable bridge linking farmers, researchers, industry, and policy sites by broadly delivering research achievements, industry trends, and outstanding field cases.
I also send strong support to all readers and to Korea’s future tech farmers. Agriculture of the future will not be an industry that depends only on experience, but a high-value-added industry that requires data and technology, environment and sustainability, and creative problem-solving capabilities together. Rather than fearing new technologies, I hope you will actively learn them, apply them to fit the field, and nurture the potential for innovation even through failure. The National Institute of Horticultural and Herbal Science will also do its utmost in developing practical technologies and supporting the field."