Japan's Smart Agriculture and Fisheries

A drone over some farmland

   Agriculture and fisheries are two industries where people have to work alongside the natural environment, which they have no way of controlling, and as both are also primary industries, they involve a lot of manual labor, leading to problems such as labor shortages and an aging workforce. People around the world are working to solve these problems using technologies like IoT and AI, and their progress is picking up speed. In this article, we will discuss some new technologies developed by smaller startups, how they are starting to see use in Japanese agriculture and fisheries, and how they may be used far and wide in the near future.

   Primary industries suffer from a variety of different issues from country to country, but in Japan these issues center on the working population growing older without enough young people to replace them. Agriculture and fisheries are known to involve a lot of hard manual labor, so technologies that reduce labor and increase efficiency are a lifeline for workers in these industries.

Predicting Harvests with Drone Imagery of Fields

The "IROHA" AI leaf color analysis service can confirm details down to resolution of just 1 mm. (Photo courtesy of SkymatiX, Inc.)

   One such technology uses AI to analyze images taken by drones, giving farmers estimates of when they should harvest and what kinds of yields to expect — all without having to visit the field. With high-resolution imagery, it can even accurately count the number of fruits or vegetables in a field, as well as ascertaining their size and how well they are growing.

   High-resolution images also let the AI detect leaf color and catch any weeds that may have started growing. This lets farmers promptly find and fix problems such as pests, weeds, and plant diseases, making it possible to maintain crop health more efficiently. The more farmers use these services the more data they accumulate, which allows them to compare the state of their fields to previous years. Agricultural land and crop management can now be carried out using a computer, instead of relying on experience and intuition.

Helping Farmers with Affordable Robotic Harvesters That Are Simple to Set Up

   Companies have developed affordable robots that can harvest crops in place of humans. For example, one company manufactures a device which is suspended on rails in greenhouses and is capable of harvesting bell peppers, cucumbers, and other produce. It doesn't have any issues with rough terrain or obstacles, so it can harvest reliably and efficiently. It's a simple robot with only the necessary features, earning it praise as a cheaper and more accessible alternative to conventional harvesters.

This robot is specialized for moving around greenhouses and harvesting fruit and vegetables such as bell peppers, cucumbers, and tomatoes. (Photo courtesy of AGRIST Inc.)

It moves around suspended on rails, eliminating any problems with rough terrain or obstacles. (Photo courtesy of AGRIST Inc.)

   In addition to harvesting, the robot can be equipped with a camera to collect images for analysis, allowing farmers to monitor their crops' health. This machine lets farmers harvest at the perfect time, allowing them to reap bigger harvests from the same area. It also addresses labor shortages, and can help keep down staffing costs.

Using IoT and AI to Reduce the Cost of Land-Based Fish Farming

   The fisheries industry also has its own high-tech startups, including one launched out of Kyoto University and Kindai University. After extensive research, they succeeded in increasing the edible portions of madai (red sea bream) by 20%, while using 20% less feed.

"22nd century tai" with more meat (top) and madai farmed normally (bottom). (Photo courtesy of Regional Fish Institute, Ltd.)

   The company is currently researching practical applications of developing fish optimized for rearing environments and raising them in land-based fish farms. They can use AI and IoT technology to automatically monitor changes in fish growth, water temperature, oxygen levels, etc., and adjust the environment accordingly — it can even be used to automate processes such as feeding and tank cleaning. They even intend to develop a device that saves on energy by using filtration technology and clever tank design to a system that naturally maintains the ideal environment for its fish.

Research is ongoing to find the ideal fish for land-based aquaculture, while also developing systems optimized for land-based fish farming. (Photo courtesy of Regional Fish Institute, Ltd.)

   This low-cost, land-based aquaculture system has attracted attention worldwide for its ability to offer reliable and relatively affordable fish even in regions with limited access to the ocean or rivers.

Preventing Pollution with Smart Feeders for Marine Fish Farms

   Marine fish farms have had issues with pressure from soaring feed prices and labor shortages, as well as marine pollution from excess feed escaping into the ecosystem. One startup in the fisheries industry has developed AI-equipped smart feeders to address these issues. The system features a camera and AI and is powered by a solar panel.

The "UMITRON CELL" gives fish the right amount of feed at the right time thanks to its camera and AI. (Photo courtesy of UMITRON K.K.)

   This system can identify three levels of hunger in fish based on the way they swim, and will reduce or adjust the amount of feed provided to fish when the AI determines that they aren't hungry. The fish can also be fed remotely using real-time video on a smartphone or other device. This reduces how often farmers have to visit the tanks, increasing efficiency.

An AI determines when fish are hungry from the way they swim. (Photo courtesy of UMITRON K.K.)

   On top of that, the system also reduces the cost of raising fish by reducing the amount of feed wasted, as well as preventing pollution resulting from excess feed. For example, in a proof of concept for madai aquaculture they successfully reduced the amount of feed used by 20%, while reducing the rearing period from one year to just 10 months. They also found that it could reduce the amount of greenhouse gases generated by wasted feed by about 20% compared to a conventional automated feeder.

Fish farmed using this system. (Photo courtesy of UMITRON K.K.)

   These AI-equipped smart feeders are also used to farm Donaldson trout at Lake Titicaca in Peru. Japanese technology is frequently used to resolve issues in primary industries around the world, such as improving working environments and building sustainable production systems.