Tokyo and Fukuoka, Japan, April 12, 2018

Kyushu University and Fujitsu Limited today announced that they will conduct joint research in the field of agriculture, for two years beginning April 2018, with the goal of increasing the stability and sophistication of agricultural production using artificial intelligence.

In this joint research, Fujitsu and Kyushu University will take plant growth status data, including plant height and the leaf surface area measured using Kyushu University's phytometric (1) technology, and a proprietary plant growth model that integrates the spatiotemporal change data (2) of plant theory, and integrate with a Fujitsu-developed AI engine able to predict information including growth rate and harvest periods, in real time.

The two organizations will also conduct R&D on systems that can bring about efficient plant cultivation, in line with demand, by regulating plant growth in the field based on these predictions.

Fujitsu will leverage the findings of this joint research in the Fujitsu Group's agriculture businesses, and together with an evaluation of effectiveness, it also aims to deliver new technologies as solutions to the agriculture industry. Kyushu University will offer education in smart agriculture, promoting dissemination of these results, together with human resource development.

Background

The stable supply of plant products and the improvement of revenue remain important issues in the field of agriculture.

Various factors, however, including good or bad weather and plant growing conditions at typical agricultural production sites, can make it difficult to flexibly supply such products that meet demand. Until now, know-how related to plant cultivation has often relied on the expertise of individual farmers, creating wide variability in the timing of the harvest and in yields and quality between individual farm fields. It has become essential to improve on-the-ground agricultural production in order to achieve consistency in terms of timing, yields, quality, and pricing.

Kyushu University and Fujitsu have been conducting joint research in the field of mathematics aimed at resolving issues in society since 2014. Now in new research, based in part using the results of earlier efforts, they are pursuing more stable and advanced agricultural production.

Summary of the Joint Research

Throughout the course of this joint research, Fujitsu and Kyushu University will incorporate Fujitsu's image processing technology, which features high recognition accuracy, into Kyushu University's alternative plant phytometric and assessment technology. In so doing, they will develop new technology that collects image data on crops using cameras and automatically measures different aspects showing the growth status of plants, including height, number of leaves, inter-nodal length, and stem diameter. Using Fujitsu's AI engine, which will combine this data with Kyushu University's plant growth model, this research additionally aims to make the optimal environmental conditions to ensure crop quality and the target harvest date, creating a system that regulates the environment according to crop growth status.

Figure: Diagram of the joint research project (credits: Fujitsu and Kyushu University)

1. Research period

April 2018-March 2020

2. Location

A smart greenhouse located on Kyushu University's Ito Campus, Fukuoka Japan

3. Roles and Responsibilities

Kyushu University

 

  • Environment for plant growth as well as R&D, provision of phytometric data measurement technology
  • Conduct research aimed at applying the plant growth model to an actual environment
  • Conduct research and development aimed at applying technology that provides timely control of the local environment (3) to an actual environment
  • Develop and implement a smart agriculture education program

Fujitsu

 

  • Integrate the plant growth model into an AI engine
  • Plant feature quantity extraction and growth estimates using image processing technology as part of phytometric measurements
  • Predict plant growth speeds and harvest period using the AI engine and optimize environmental control
  • Create a system supporting plant cultivation that integrate the technological elements listed above

Future Developments

Using the results of this joint research, Fujitsu aims to deploy this technology in farm fields operated by Fujitsu Group companies in the agriculture business, and to make it commercially available as a solution aimed at the agriculture industry. In order to promote the use of ICT in the field of agriculture, Kyushu University will work to bring about smart agriculture, not only within Japan but also in the Asia region and other countries and will actively contribute to the nurturing of human resources.

 

  • [1] Phytometric

    Image diagnostic technology applied to leaf surface area or plant height. It also includes proprietary technology for the measurement of photosynthesis, evaporation, water absorption by roots, and conveyance of photosynthetic products (sugars) to fruits.

  • [2] Spatiotemporal change data

    Information on changes due to time and position or location. This case refers to information related to growth, such as photosynthesis and translocation within the plants.

  • [3] Technology that provides timely control of the local environment

    Technology to control the environment at a precise level, such as controlling the temperature of only a specific part of the plant for an appropriate length of time

Source: Fujitsu

Comments

No comments to display.