Farmer A, a young agriculturalist who discovered suspected pest symptoms in a rice paddy, previously had to ask village elders or search separately through multiple systems for diagnosis, soil, weather, and pesticide information. Now, however, farmers can simply take a picture of their rice with a smartphone and ask "Isaki," an agricultural information AI service developed by the Rural Development Administration (RDA), to instantly identify pests and find solutions. Furthermore, the RDA plans to introduce "Saessaki" to support researchers in their research, technology dissemination, and administrative tasks.

According to the RDA on the 20th, downloads of Isaki reached 17,819 and agricultural inquiries hit 90,000 between January and August of this year. An RDA official explained, "AI Isaki was created so that farmers do not have to jump across multiple systems, but instead get necessary farming information by conversing with a single AI." They added, "By building upon 560,000 pages of specialized agricultural knowledge and 300,000 pest images accumulated by the RDA over a long period, we have secured the professionalism and reliability of its answers."

Ask and Snap, and AI Answers... The "Doctor-Level Farming Assistant" in Farmers' Hands

On the 30th of last month, the RDA released "Isaki 2.0," featuring enhanced response performance and field usability. The most significant update is the video/image diagnosis function that identifies pests through photos. Additional features include fertilizer recommendation and voice farming diary functions, alongside an optimization for mobile environments. If a case is not pest-related, the system searches registered professional books (Agricultural Encyclopedia) to provide similar symptoms and answers. A quality assurance team—consisting of dedicated personnel and 25 technical committee members across various fields—is also in operation.

Work is also underway to connect scattered information systems into a single hub. Plans are in place to link 36 services across 17 information systems related to pesticides, weather, fruit growing, food, business management, and livestock. Once integration is complete, users will be able to check weather, soil, pest risks, and pesticide registration information all at once with a single question. Next year, the RDA plans to upgrade the service to "AI Isaki 3.0," adding customized management consulting, data linkage with external institutions, and web search capabilities.

Learn, Share, and Build: Staff Members Use "AI Saessaki" for Work

While farmers have Isaki, RDA staff members have Saessaki—an operational AI supporting research, technology dissemination, and administration. A representative prototype, the "AI Fellow Researcher," connects 12,504 past research data over the last 10 years to review new projects for redundancy. The National Institute of Agricultural Sciences built a system to bundle fragmented technologies and discover field-demonstration research, while the National Institute of Crop Science established an administrative support chatbot to assist with complaints, research project information, and National Assembly responses.

The people who will use AI are also being nurtured together. The RDA is recruiting 13 experts in cutting-edge fields such as digital biology, AI model development, and agricultural robotics. It is also promoting practical AI training matched with job duties by classification, alongside evaluations of employee AI and data competencies.

The RDA is not stopping at merely "using" AI. A total of 1,106 people participate in the online knowledge-sharing community "AI Dure Turubat" to share utilization experiences. In the "Together AI Ideathon," where staff design the AI they need, 8 out of 59 ideas were selected as winning entries. Tools such as "Chajeumi" (recommending responsible persons based on work areas), "Gwasu-jikimi" (decision-making support for abnormal weather in orchards), and the "Korean Report Assistant" originated from inconveniences staff experienced while working. Through this structure, the RDA discovers field demand via ideathons, verifies usability through prototypes, and infuses generally useful functions into AI Saessaki to spread them across the entire organization.

Accelerating Field Delivery through "Data Opening and Research-Dissemination Linkage"

The core of AI services and research innovation ultimately lies in data. Accordingly, on July 14, the RDA launched a national key data-opening project to establish a collection and opening system for AI training pest images, while expanding collection scopes to unstructured data such as photos and videos. The "Agricultural & Rural Public Data + AI Utilization Startup Contest," co-hosted with the Ministry of Agriculture, Food and Rural Affairs, drew 290 teams, with 10 award-winning teams utilizing RDA public data.

Changes are also being pursued away from the conventional "research-first, dissemination-later" approach. Moving past distributing technology only after research concludes, the RDA is building a converged business framework where dissemination departments participate from the planning stage. It also plans to mandate "field performance evaluations" before research projects end so that developed technologies can be rapidly distributed to the field. "Advanced technology scouting"—a trial program bringing in necessary cutting-edge technologies from the outside—is also being test-run. While checking the field adaptability of pilot projects such as autonomous transport robots, the RDA is cooperating with the Korea Technology Finance Corporation and the National Testbed (NTB) to quickly discover and match private technologies applicable to agriculture.

RDA Administrator Lee Seung-don stated, "Through AI Isaki, farmers can now receive the RDA’s latest agricultural technology information in real-time, free from time and place constraints. Through continuous upgrades, we will develop it into an intelligent farming assistant that all farmers seek out in the agricultural field." He added, "With 'AI Saessaki' built directly by our staff, we will transform how work is done across research, technology dissemination, and administration, ensuring research outcomes reach the agricultural field even faster."

[Copyright (c) Global Economic Times. All Rights Reserved.]