BreakThrough Study Idea

Agriculture is unmanned and becomes a tourism resourceMasayuki Hirafuji<Series 3 / complete>

Study Idea


  • facebook
  • twitter

Acquiring big data will create new business opportunities

The number of agricultural researchers is decreasing in the long run. About 30 years ago when I entered the Ministry of Agriculture and Forestry, agrarian researchers of Japan where thousands of people have been reduced to 1,400 people now. Also, the number of agricultural researchers in prefectures is decreasing at the same time. In other words, even in research on agriculture people can not mobilize. If it is not automated any longer, agriculture tapering will be inevitable.

Currently, I am engaged in fundamental research to collect big data in JST CREST research project. Tokyo University, Agricultural Research Organization, Tsukuba University, NEC Solution Innovator jointly integrate data acquired using field servers, drone, and a six-legged robot, aiming to discover new findings by making it big data. In this essential research, we are trying to elucidate the mystery of hybrid vigor and the magic of symbiotic microorganisms by acquiring time series data such as crop yield, plant height, growth rate, environment and so on. If you discover the mystery of symbiotic organisms affecting growth, disease resistance, aroma, taste, freshness, etc., new business opportunities will be born. However, it is difficult to acquire the data of growth accurately. Data on plants and the environment must continue to be collected at many points for a long time. Installing an excellent field server near crops will interfere with machine work such as pesticide spraying and weeding. When trying to photograph crop seedlings with millimeter precision with a drone, seedlings will also move due to the wind blowing when the drones fly, besides the wind blowing naturally, so it will not be able to measure accurately. If cooperative flight with multiple drones is realized, that problem may also be solved.

Members of field server, drone, data collection experiment with a six-legged robot

In recent years, it has become possible to measure micro traits by attaching a microscope camera to a 6-legged walking robot. You can automate the sampling of crops that you have done manually. From the image, machine learning such as deep learning, etc., it became clear that the flowering time of plants and the number of spikes were also visible. Robots and AI make it easy to acquire data automatically, but there are issues as well. That is what we have to teach every single meaning such as "This is rice plant," or "This is rice" and that the hand of man must measure the actual value data for that. To obtain accurate value data, a massive amount of sample is necessary, so it takes much effort to get distracted.

What is the form of next-generation agriculture represented by data?

So, the challenge is "Farm producing true value data." It is an effort to automate the acquisition of true value data by creating "Data farm" which is a field for mass-producing data utilizing sensors, cameras, image processing, etc. in Sarabetsu village, Hokkaido. In recent years, there are many people with high IT skills in Tochigi, such as large-scale management using free running agricultural machines using RTK-GPS which can perform positioning with an accuracy of a maximum error of 3 cm, and the number of people who are taking master's degree and a doctorate in agriculture is increasing. Some farmers can do all of the experiment field management, drone maneuvering, and data management in Sarabetsu village, and ordered all of the field construction, agricultural work, drone shooting, maintenance, data management and so on. As these farmers increase in the future, it will be possible not only to produce crops but also to produce big data necessary for machine learning and research development. In the "Data farm" of Sarabetsu village, we produced data and constructed a system to harvest data with field server and drone. The harvested data will be useful for breed improvement, pesticide development, cultivation technology development, etc. An offer has already come from a pesticide manufacturer etc.

Agriculture will become unmanned in the future. In Star Trek's worldview, farming was a hobby, but it will be exactly the way it is. If so, state-of-the-art agriculture will also be a vast tourism resource that provides a science dream aimed at inbound. Indeed, in 'Epcot' which is one of the Disney World of America, attractions based on the future agriculture theme are becoming popular. Although agriculture has a significant initial investment, if investment resources can also be utilized, investment momentum will increase in the future.

Next time I will explain the possibilities of agriculture that can change the world.

< series 3 / complete >

Masayuki Hirafuji
University of Tokyo graduate school Agricultural Life Science Graduate School attached Ecological Harmonization Agriculture Organization International Field Phenomenic Research Base Appointed Professor Dream Science Holdings Co., Ltd. · CEO

1983 Graduated from the University of Tokyo Life Science course. He experiences such as Ministry of Agriculture, Forestry and Fisheries · Agricultural Research Center (Current Agricultural Research Organization) Chief Researcher, Research Agency for Agriculture, Forestry and Fisheries Technology Conference Secretariat, and Director of Agricultural Research Institute so far. He specializes in computational biology, modeling and measurement control of biological and ecological systems, research on space farms, comfortable living spaces, using IT, complex systems, artificial ecosystem technology.

◇Main book
"Bioexpert systems - AI/neurocomputing in biological production" (Corona Inc.) published in April 1990

share by SNS

related article