Shizuoka University - AI to support agricultural production

Country
Japan
Sustainable development goals
  • Decent work and economic growth
  • Industry innovation and infrastructure
Project link
https://business.nikkeibp.co.jp/atclh/NBO/mirakoto/mirai/073000027/

About the project

Shizuoka University is working on a project that uses artificial intelligence to support agricultural production by automating certain processes involved. The system aims to inform farmers of the appropriate time to water tomatoes and to assist in the sorting of cucumbers in the pre-sale process.

More project information

The project uses AI system to support agricultural production by automating certain processes involved. The system aims to inform farmers of the appropriate time to water tomatoes and to assist in the sorting of cucumbers in the pre-sale process. The project is currently in progress, aiming for wider use of artificial intelligence for more varieties of plants and purposes.

July 2017, the project succeeded in verifying the usefulness of the system in tomato production. Through cameras and sensors set in greenhouses, the artificial intelligence system monitors the thickness of the stem, a factor that indicates appropriate timings to water the tomatoes. The system can analyze the images taken every few minutes of the plant to sense subtle changes that only experienced workers can recognize. It can also learn the relation between the thickness of the stem and other factors in the environment such as humidity. 

Shizuoka University also started working on ways to produce sweeter tomatoes by appropriately limiting the amount of water from observing the extent of withering of the leaves using artificial intelligence. By putting stress on the vegetables through the reduction of water supply, farmers are able to produce sweeter tomatoes. However, this method had been very difficult and required the instincts and decision-making of skilled workers since limiting water supply excessively or inappropriately can lead to loss of crops. The project aims to quantify subtle changes and details of plants into objective data to allow high quality production of plants without relying on the skills of highly experienced workers. Shizuoka University will continue with the project to make the artificial intelligence system applicable to other varieties of plants. In addition, Shizuoka University developed an artificial intelligence that automatizes the process of pre-sale selection of crops, specifically cucumbers, using an image recognition system that distinguishes different sizes and conditions.

The project contributes to the goal of Industry, Innovation and Infrastructure, as well as Decent Work and Economic Growth by automatizing certain processes in agricultural production with the use of artificial intelligence and allowing more human labor to be used for advancing other areas. The project helps the current agriculture industry in Japan that faces serious problems in terms of labor shortage and the aging of the work force.