AI Pulls Tractor... Uses Deep Learning Technology to Break Soil and Optimize Driving

Establishing Industrial Foundation for Autonomous Driving Tractors
Nongjincheong: "The Era of Unmanned Agriculture is Near"

[Image source=Yonhap News]

[Sejong=Asia Economy Reporter Moon Chaeseok] A technology has been developed in which artificial intelligence (AI) autonomously recognizes soil clumps and other obstacles that interfere with driving a cultivator and assists in adjusting the driving path.

The Rural Development Administration announced on the 4th that it has developed autonomous driving technology for tractors.

The system captures images of dirt roads with a camera installed on the tractor and runs deep learning technology. The AI analyzes how much soil breaking work blocking the cultivator's path has progressed and adjusts the driving direction by the error from the driving baseline. Using this technology, when autonomous driving was performed straight at a speed of 3 km/h, the resulting error was within ±9.5 cm. This error was more than twice smaller than the ±21.2 cm error when driven manually by a driver.

The technology development is expected to enable cost reduction. This is because it only requires processing AI’s visual information without the need to use existing satellite navigation systems. It has also become easier to instantly recognize obstacles and respond to unexpected situations.

The technology has already been filed for industrial property rights, and related companies plan to enhance its completeness further before practical application.

Additionally, the Rural Development Administration has developed advanced tractor autonomous driving technology using a high-precision positioning system. The positioning system refers to location determination technology that allows the tractor to recognize its current position, speed, and route. As the precision of the positioning system improves, it becomes possible to preset the work path and perform straight driving, turning, and other maneuvers according to the situation.

Agricultural machinery companies plan to apply the straight autonomous driving technology to tractors, distribute it in rural areas, and industrialize advanced technologies such as turning, obstacle recognition, and avoidance.

Kang Geumchun, head of the Smart Farm Development Division at the Rural Development Administration, said, "We plan to establish a cooperation system among industry, government, and research institutes to ensure that the technology developed by the Rural Development Administration can be quickly applied in rural fields, and to prepare policies, support projects, and certification standards early." He added, "Soon after industrialization, the era of unmanned agriculture will become a reality."