AgriTech: IoT and the Future of Food Security
Although spending is predicted to double on IoT ventures in the next four years, few companies are directing their attention to the agricultural applications of IoT. As the population of the earth exceeds our potential for food production, technology companies and farmers are looking for ways to collaborate. Their goal: To extract the highest quality and quantity of crops from the least amount of arable land. These partnerships will usher in major changes to farming as we know it – expedited by climate change and the consequent unpredictability of weather patterns. Soon, more and more farmers will shift their operations into the more easily controlled environment of greenhouses.
This movement is not limited to large scale agri-businesses and corporate giants. Partnerships between technology companies and agriculture are possible on all scales due to the affordability of IoT devices like sensors and Bluetooth gateways. Sensors help growers monitor soil quality, microclimate conditions, groundwater, inventory traceability, and building management. With sensors deployed throughout fields and greenhouses, farmers are able to make smart decisions concerning fertilization, spraying, and irrigation based on real-time sensor data.
The following are a few ways in which IoT is set to disrupt agriculture and provide hope for a future of food security.
Phenotyping involves the selective breeding of plants to produce strains which are more resilient to changes in soil quality, weather, and irrigation. By placing sensors which detect soil pH and moisture content, farmers can create a 3-D image of a field and determine which plants fared best, even in poor conditions. Technology takes much of the guesswork out of selective breeding. With accurate data on soil uniformity, moisture levels, plant height and density, farmers can breed pest, frost, and drought resistant strains of crops.
Early Detection and Disease Prevention
Technology is making it easier for farmers to detect the early stages of pest infestation and disease in all manner of crops via imaging, machine learning, and light wavelength analysis. In developed countries, less than 60% of production escapes the ravages of pests, pathogens, weeds, and insects. The social cost of pesticides make them an unsuitable solution – for every $1 spent on pesticides, the damage to the environment and human health is estimated to range between $5 and $10 dollars. Mass spraying is no longer a sustainable or viable option for farmers in developed countries. Companies like Monsanto which produce weed-killer are under fire for directly contributing to human instances of cancer. Many other pesticide producing companies are under scrutiny for their role in the global crash of the pollinator population.
For these and other reasons, there is great pressure from governments and the public to find other solutions to manage pests and diseases. Agri-giants have already adopted high tech solutions. One method of prevention is imaging. It has applications in-field and indoors.
In greenhouses where plants grow on fixed tables, an extremely powerful camera can move between rows of plants, taking images of plants and sending those images to a processor to assess for signs of mildew, mold, or pests. The software used in such cases is trained to separate image foregrounds from backgrounds and to detect anomalies with a high degree of accuracy. Imaging and machine learning are also used in outdoor farming operations. Drones, equipped with infrared cameras, can snap aerial shots of fields to detect land gradation and general plant health.
Farmers are also using light sensors to detect disease and pest infestation. All plants absorb a small percent of sunlight during photosynthesis, while the rest is reflected off the plant’s leaves. By detecting how much light a plant is absorbing, software programs can tell farmers how much stress a plant is under.
Light sensors can also help high tech sprayers detect which plants are crop and which are weeds. “Smart” Herbicide Sprayers are equipped with chlorophyll-detecting fluorescence sensors. Each sensor takes a scan of the ground in 4 sections, analyzes the data with edge computing, and directs the spray nozzles where to place a sub-second stream of herbicide. This targeted spraying lets farmers increase or decrease the volume of herbicide used depending on how thick an area is with target weeds.
“They shine a red light on the plant, and then the wavelength of that light gets shifted in the chlorophyll of the plant and a little tiny amount of it gets beamed back out of the plant as near infrared,” said Adam Huttton of WEEDit. Instead of blanket spraying an entire field with herbicides, spot targeting severely reduces the risk on environmental contamination and pollution from excess spray.
More heads are better than one when it comes to making weather, seed, and irrigation related decisions. The mass of sensor data collected from one growing operation can be aggregated with data from other networks to create a more complete and useful map of regional conditions. These “agronomic” systems aggregate and process data – providing farmers with actionable insights on seed choices, soil qualities, and rainfall.
In the U.S, the Farmer Business Network (FBN) connects upwards of two thousand farmers in 28 states with data regarding coffee prices, profitability, and agronomic performance.
“Farmers can see how seeding dates, population density, precipitation levels, rotations, and soil temperatures affect yields in their area.” (Sean Pratt, 2016)
Greenhouses provide the ultimate controlled environment for farmers and hobbyists to perfect conditions. Light levels, soil quality, irrigation, humidity, and temperature can all be micro-monitored and tweaked according to the needs of separate plant beds. One company in California has taken this to an extreme with a fully automated operation in a warehouse.
Iron Ox Hydroponic Farms CEO, Brandon Alexander, stated “If we’re going to need to double the food production in the next 30 years to feed the growing population, we felt there needs to be a radical change.”
The company uses a modified warehouse as their growing space and two one-ton robots which patrol the facility on wheels through wide aisles, imaging, watering, shifting, and kerning individual plants to optimize production. The robots sort the plants by their life-cycle stage to make the best use of space within the greenhouse. Young plants are spaced closer and mature plants are spaced farther apart. This practice optimizes plant density and the distribution of light and fertilizer.
Iron Ox is able to to the equivalent of 30 acres of outdoor farming in just a single acre of our robotic farm. The company aims to ensure that urban communities have access to healthy, affordable food. By growing food locally, the company is eliminating the carbon footprint which results from shipping crops (on-average) 2,000 miles. Cutting transit costs and automating labour can ensure that produce remains affordable.
As the population increases, cities and suburban sprawl are consuming valuable farm land – contributing to unsustainable erosion. According to David Pimental, “It takes approximately 500 years to replace 25 millimeters (1 inch) of topsoil lost to erosion. The minimal soil depth for agricultural production is 150 millimeters. Greenhouses and environmental monitoring devices are likely to shift from novelty farming methods to the standard, both out of a desire to prevent further land consumption and from necessity – the quality and quantity of land is diminishing unabated.