"The reason Daedong is creating artificial intelligence (AI) is clear. It is to incorporate AI into products and services so that farmers can achieve greater income. We will lead the transformation of Korean agriculture, which is at a turning point, by applying AI in the three key areas of movement, operation, and cultivation."
On the 17th, CEO Choi Jun-ki of Daedong's AI robot software subsidiary, Daedong AI Lab, noted at a media briefing held at the Daedong Seoul office in Seocho-gu, Seoul, that this is the reason Daedong is developing AI.
Previously, Daedong established Daedong AI Lab in May 2024 and has been focusing on AI development with an annual expense of about 3 billion to 4 billion won for research and development (R&D). The briefing was organized to share the achievements of the past year and explain the direction of the development.
CEO Choi especially said, "We will create an easy and smart agriculture," adding, "We plan to develop AI centered around three pillars: autonomous driving, agricultural work, and cultivation to apply it to all group businesses, including smart agricultural machinery, agricultural robots, precision agriculture, and smart farms."
Daedong will first develop movement AI to support agricultural machinery and robots in independently determining paths and performing optimal tasks. CEO Choi said, "The AI for tractors has unique characteristics different from other fields, such as automobiles," and noted, "We have worked hard to develop technology that is more precise even if it is slower, with an error range of less than 7 centimeters."
CEO Choi continued to explain that agricultural machinery used in farmlands has also enhanced capabilities for recognizing the boundaries of fields and responding to obstacles by equipping "vision-based autonomous driving technology." To accomplish this, Daedong has collected about 500,000 images of orchards and fields and approximately 3 million driving videos since last year, securing the largest agricultural data in the country.
Daedong AI Lab plans to apply AI-based algorithms to group products such as the RT-100 in the second half of 2026, starting with the flagship tractor, which is in the fourth stage of autonomous driving. It will also establish a machine learning operations (MLOps) system to continuously reflect customer operation data in learning to enhance performance.
CEO Choi also stated, "Most agriculture relies on manual labor and faces difficulties," adding, "We will automate hard manual work and enhance agricultural efficiency through 'work AI' technology." Work AI is a physical AI that performs various agricultural tasks such as tilling, sowing, fertilizing, pest control, and harvesting using agricultural machinery and robots.
First, we are focusing on the representative task of tilling in tractors. We are developing an autonomous working model that analyzes soil conditions through cameras and sensors and evaluates work quality in real-time. This model aims for up to a 30% reduction in working time and a 15% improvement in fuel efficiency compared to existing autonomous work.
Furthermore, we have begun enhancing the multifunctional robot based on imitation learning disclosed at CES 2025 to assist in cultivating strawberries and tomatoes. However, he stated, "Currently, it is difficult to apply in the field due to a lack of data recording or capturing physical phenomena," and added, "We are at the stage of collecting remote operation data for various crops to train robotic arms or hands."
Finally, CEO Choi noted, "It is important to foster smart farming to overcome the decline in agricultural production due to climate change," and stated that they are developing cultivation AI. To this end, Daedong predicts crop growth based on data collected from satellites, drones, and smart agricultural machinery and provides tailored cultivation strategies. Notably, Daedong AI Lab plans to prioritize the development of AI needed for strawberry cultivation and then expand its application to various crops.
During the following briefing, when reporters asked about 'government policies needed during the agricultural transition', Daedong Executive Director Na Young-jung said, "If we can collect crop or work data in the agricultural sector through the government, it will help the development of agricultural AI," and added, "We will focus on creating services that are helpful in the private sector based on various data."