Agriculture Industry Moves Forward Using Artificial Intelligence (AI) To Improve Crop Management
28 de maio de 2020It is always fun to look at the widening expansion of sectors that are being helped by artificial intelligence (AI). Farming has regularly used technology to improve yields
By David A. Teich
In recent years, global warming has made it more important to manage water resources through improved irrigation. Now the agriculture industry is looking at adopting AI in many ways. One of those methods is to analyze crops to better manage yield.
Aerial imaging for crops is not new. Frederick Bawden used it in 1933 to detect diseases in potato crops. The technology has advanced since then and now AI is being introduced to move forward yet another step.
As an example, imagine an irrigation system that uses center pivot irrigation. While farmers are moving to other, more water efficient systems, that is expensive and takes time. In the meantime, it’s important to quickly find problems with the irrigation system to control costs and protect crops.
Ceres Imaging is a young company working to improve crop management. Their team uses AI in multiple ways. That begins in their scanning technology, using advancement in vision to do aerial inspections of fields via equipment mounted in planes. They first considered drones, but those have limited flight times and carry weight. An airplane is much more efficient, covering more acreage at lower cost.
Their vision technology can detect many problems in fields. In the above example, when that system component identifies a circular area that is drier or wetter than the rest of the field, the machine learning (ML) component can recognize a problem with a central pivot and the overall system can notify the farmer.
Other problems, involving nutrients, pests, and more can be identified by the ML system. As you have probably already guessed, supervised learning is used to train the ML system. There are many images and data sets than can be annotated for training. It is not only historic information, as millions of acres are being imaged every year, often multiple times.
“Farming is a low margin business and you need the full picture to optimize every penny you’re putting into your operation,” says Ashwin Madgavkar, CEO, Ceres Imaging. “Getting ahead of issues using artificial intelligence is worth its weight in gold, allowing for rapid correction of problems before they impact yield.”
It is also important to know that the technology isn’t limited to crops. Orchards are a similar issue, and already a market. The basics of irrigation, pest control, and more are similar to crops and enough labeled data is available to train the ML system on those products.
Mr. Madgavkar says twenty percent of California’s vineyards are currently being observed via their AI system. I can see the opportunity for other sectors, such as forestry for paper, but Ceres Imaging is focusing on the existing markets while they grow.
Farmers, Brokerages, Crop Insurance All Benefit From Timely Information
Farming has many variables even in more stable times. Global warming is making it even more uncertain, as can variable labor risks – as the current pandemic is showing. In the agriculture industry, the farmers aren’t the only people whose success is closely linked to crop yield. Crop insurance and the brokerage industries exist by managing risk in order to profit and succeed. Companies in those areas also need to know what is happening in fields.
Artificial intelligence that provides better information about crop yields helps those to ag sub-sectors to better plan for their own needs and solvency. That information then can indirectly help the entire food chain, helping manufacturers better plan for expected supply of produce.
As global warming and the Covid-19 pandemic are showing us, too many people in the advanced world have taken the food supply chain for granted. It is more fragile than most people realize, and AI can have a powerful impact in mitigating the risk – starting at the source.
HTTP Status Codes Command This Malware How to Control Hacked Systems
Blockchain and Decentralized Identity: How Transparency Can Improve Identity Verification for Banks