Full-Scale Rollout of AI-Driven Innovation on Farms

Sejong, 6 May 2026 — The Ministry of Agriculture, Food and Rural Affairs (MAFRA) has announced that it will provide further momentum to the promotion of artificial intelligence (AI)-based smart agriculture by utilising 32 graphics processing units (GPUs) provided by the Ministry of Science and ICT’s ‘National AI Project for Countrywide Artificial Intelligence-Driven Innovation.’ 

The Ministry plans to utilise its 32 allocated GPUs in collaboration with the private sector. First, the Ministry will employ 16 of the GPUs to analyse crop growth images and video files through GPU-based AI models and build datasets for training AI models on crop growth. This is expected to improve the quality of smart farm data and its applicability to the AI transformation effort by automatically extracting key information on crop growth status that was previously measured manually.

The remaining 16 GPUs will be provided to enterprises developing AI solutions for smart agriculture applications. These businesses will use the GPUs to combine government-created smart farm data with their own data in order to build high-quality AI training datasets. Their applications will also be shared with the related industry.

In this connection, the Ministry has taken steps to better understand industry needs and refine support strategies. On 20 April, the Ministry organised a demand consultation meeting with AI solution developers in greenhouse horticulture and open-field farming as well as in advanced agricultural machinery. 

The meeting attendees presented various scenarios in which the GPU would be beneficial. For example, AI solutions that analyse greenhouse drainage solution data could be used to prescribe cultivation strategies. GPU-based models could process terrain and large-scale environmental data for open-field smart farming, and shared-use autonomous-driving AI models could be developed to realise physical AI in agricultural operations.

The participating companies said GPU support would accelerate the analysis and use of agricultural big data. This would allow them to develop advanced AI solutions previously limited by computing constraints. They also anticipated that integrating government smart farm resources with their own on-site findings would create a growing pool of datasets for AI training, which would help raise overall data quality and expand the application ecosystem.

The Ministry will reflect the findings from the demand consultation meeting in its public call, scheduled for launch in May, for companies to participate in a high-quality AI training dataset development project. The selectees will be granted use of up to four GPUs, along with access to government smart farm data. At the same time, the government will collect the datasets generated through AI solution development by the selectees and make them publicly available through the Smart Farm Big Data Integrated Platform website (https://smartfarmkorea.net), thereby expanding the foundation for data utilisation.