Four things you should know about Smart Farming
Robotractors coming to a field near you.
From watering crops to harvesting, a farm in Japan will become the first in the world to run on robots from next year. Using Smart Farming techniques, the robots will complete almost every task on the farm, doubling its daily output.
GovInsider spoke to Dr Masayuki Hirafuji, Director of Large-Scale Farming Research Division, at Japan’s National Agriculture and Food Research Organisation (NARO) to find out what Smart Farming means for governments.
What is Smart Farming?
Smart Farming uses sensors and big data to more accurately predict and make decisions to make farming more productive.
“With the evolution of sensors and communication technologies, we will be able to have more and different kinds of data which could save us lots of money,” he said.
In the future, this would mean using more artificial intelligence for faster and more accurate calculations and using more robots to replace manual labor.
Why is it important?
Farmers still make important decisions, like how much pesticide to use or which kind of seed to use, based on intuition, rather than precise mathematics. “Intuition is not enough to optimize farming, so farmers need more accurate predictions for their decision-making,“ Dr. Hirafuji said.
In Hokkaido, Japan, for instance, sensors tell farmers how much pesticide they should use. “It has an infrared sensor so it can measure the amount of nitrogen in the soil, and a computer can estimate the requirement of fertilizer in each area,” he said.
Smart Farming techniques can also help researchers better understand biological mechanisms. For example, scientists know little about how breeding two plants with bad yield can sometimes produce seeds with better yield - a phenomenon called hybrid vigor. Sensors in farms can give researchers data to solve such mysteries, Dr Hirafuji said.
How does it affect government?
Smart Farming can give governments more accurate data on food security. “Satellite data can estimate the distribution of yield in the world so we can predict supply of food,” Dr Hirafuji said. Or drones with infrared light can detect crops infected by diseases, he added.
Countries where economies are driven by agriculture or where the population is aging can use robotic farms can improve productivity.
“It can improve efficiency and increase the yield of crops, and it can save the cost of labor,” he said.
Driverless tractors that navigate themselves through farms spraying pesticides or transporting produce are already available. They use the same kinds of computer programs as driverless cars, he said.
“In 30 years, computers will be smarter than us. We can imagine that only robots will cultivate farms,” he said.
Countries where smart farming could have the biggest opportunities also face the biggest obstacles. The ones that are economically dependent on agriculture also tend to have smaller farms, making it more expensive to implement these techniques.
“Smart farming is still very expensive, and it’s useful only in large-scale agriculture,” Dr Hirafuji said. Farms should be at least 40 hectares for this to be profitable, he estimates.
“The first step is that governments should encourage large-scale farming,” he said. In Japan, with a growing number of elderly farmers retiring, farms are predicted to double in size over the next 20 years, he said.
What is happening in the region on this?
Japan will next year use artificial intelligence to more quickly find new food crops, Dr Hirafuji said. Most food is produced from plants (called cultivars) that have been deliberately bred to produce a certain yield, flavor, and resistance to diseases.
“Typically, we need 10 years to produce new cultivars. By using AI, we can accelerate the breeding by two or sometimes three times,” he said. Using huge amounts of data on plants’ genes and the kinds of characteristics they produce, computers will predict how new mutations of genes will manifest on the plant physically, he said. The government is also testing whether drones can be used to take 3D images of plants.
“A swarm of drones can simultaneously take photos, and we can use this data to construct 3D models of individual plants,” he said.
“From the images, we can extract numerical data such as the number of plants, growth rate, and kinds of diseases. This is already possible using pattern recognition, an AI technology,” he said.
Meanwhile, Malaysia wants to generate U$320 million in income using smart farming. Agriculture is a key area of focus for the country’s new National Internet of Things Strategy.
In one of the first projects, the government is testing sensors to detect the right time to pollinate oil palm flowers. While in Japan robots are already starting to take over some tasks, in less than 50 years they could be running whole farms.