7 Ways Artificial Intelligence and Machine Learning Are Changing Agriculture and Farming
When you think about artificial intelligence (AI) in agriculture, a couple applications likely come to mind, but there are a lot of ways AI and machine learning technology are intertwining themselves in agricultural and farming products. Here are some areas that have experienced recent innovations:
Chickens are usually most vocal in the morning, chicken farmers say. At night or in the evening, they’re quieter, unless they’ve just laid eggs, in which case they sing their egg songs of victory. When early in the day one chicken farmer went into a large chicken coup and the chickens were quiet, he knew something was wrong. It turns out the automatic lighting system hadn’t turned off over night, and the chickens didn’t rest well.
Today, chicken farmers don’t have to physically go into the coup in the morning to discover something’s wrong. With the right technology, they can find out sooner. They can use technology that monitors chicken talk for unusual communication patterns and even sickness. This way if something goes awry, they can find a solution sooner rather than later.
When first born, piglets are at risk of being crushed by their mothers. Six million piglets die each year by being crushed this way. That’s why one agriculturally minded technologist designed a belt for the mother sow that works to prevent piglet deaths. The belt, which wraps around the pig, mildly shocks the sow, like a dog collar, but with less intensity. It zaps her only when it identifies distressed squeals from the piglets she’s crushing. This can help the mother shift and save piglets’ lives.
Many horses die from colic, a fast-moving horse sickness that can frequently lead to death. If a horse comes down with the illness and it isn’t caught early enough, the animal can die quickly. After one horse lover’s favorite steed died from colic because no one caught the symptoms in time, she dedicated herself to developing a way to prevent horse deaths from colic and other diseases.
Alexa Anthony, CEO of Magic AI, designed an AI tool that uses computer vision to monitor horse behavior. By setting up a camera in a horse’s stall, the machine learning application monitors the horse’s behavior for unusual patterns. If the system notices the horse is acting strangely, it pings the owner. When an owner gets a ping, he or she can watch a live video feed of their horse to assess for themselves if they should take further action. This way if a horse does have colic, he or she can get assistance as soon as possible.
One Swedish beekeeper was deeply concerned about the devastation varroa mites were wreaking in beehives. This deadly bee enemy sucks the life out of their victims, eventually leading to colony crash. To find out the level of mite infestation, a beekeeper usually has to kill many bees in order to also kill and count the varroa. Unfortunately there was no good way to tell how infested a beehive was without killing the bees in the process.
As a result of this, Bjorn Lagerman built an AI-powered tool, called BeeScanner, that can identify the number of bees and the number of varroa attached to those bees, giving beekeepers a handle on how infested their hives are and how to handle that infestation. As of March 2018, the app is live on Android.
Two companies in particular come to mind for tracking bovine health, each taking a different approach. One uses cameras to identify individual cows and monitor their food and water intake, odd behaviors, and heat patterns. When the tool detects something unusual, it sends the owner a health alert. The aim is to improve herd health, in turn increasing milk production.
The other company uses sensors that monitor when cows are eating, sleeping, and drinking. With this data, the AI ferrets out unusual patterns, letting the owner know when something is awry and also predicting problems in the foreseeable future. The product can even identify cow fertility. Like the previously mentioned product, this company’s tool also pings owners when it detects unusual bovine patterns.
After acquiring a company that uses computer vision and machine learning to identify weeds, John Deere is implementing this new tech into its equipment. The company it acquired specializes in precision farming. One of its products reduces the use of herbicides by identifying individual plants against a database and only spraying identified weeds. This technique avoids spraying herbicides on the crop and surrounding soil. By using less herbicides, this saves farmers money and it reduces the effects of chemicals.
Similar to the dairy-farm market mentioned above, companies are competing to optimize produce growth in greenhouses. One concept uses installs rails on a greenhouse ceiling to mount moving cameras that continuously monitor plants for infestation and plant death. When the system discovers something, it pings the farmer. This reduces the number of employees that need to walk the greenhouses checking for plant issues.
Another approach takes plant production to a whole new level. Farmers use hydroponic growth strategies, plant stacking methods, and AI to produce 100 times more produce on the same amount of land than the traditional farming method. By stacking large trays of plants in horizontal rows that stack vertically on top of each other, farmers can produce 100 times more produce on the same land footprint than the traditional farmer would on that same parcel of land. This approach also saves 95 percent of the water used in traditional agricultural approaches.
AI and machine learning are changing the farmer’s experience, like many innovations before, and this seems to be the start of the AI applications. We’re excited to see how entrepreneurs continue to innovate in the space.
Photo by Nathan Anderson on Unsplash