GeoVisual Analytics Leverages AI for Agriculture Insights

geovisual-analytics-leverages-ai-for-agriculture-insights

According to Tractica’s research, one of the industries best positioned to leverage artificial intelligence (AI) – at least in the developed world – is agriculture. In our Artificial Intelligence for Enterprise Applications report, we forecast that spending on AI software in the agriculture industry will grow from $16.2 million to $373.7 million by 2024.

Recently we sat down with Jeffrey Orrey, CEO of GeoVisual Analytics. GeoVisual is a Boulder-based startup focused on using remote sensing and big data analytics to improve and predict crop yields, better manage croplands, and improve harvests.

The company’s analysis is based on the properties of electromagnetic waves in the near infrared (NIR) spectrum, which are invisible to the human eye. While most objects reflect a negligible amount of NIR light, actively growing plants reflect a lot (up to 6 times more than a plant’s reflectance of visible green light) and plants stressed either by disease or drought will exhibit a reduction in NIR reflectance. Adopting a color filter that shows NIR as red light, green light as blue light, and red light as green light, actively growing vegetation shows up prominently on images as bright red, stressed vegetation shows up as a darker red, and non-vegetated areas, such as lakes, show up as blue or green depending on their composition. In addition, there are subtle differences in NIR reflectance between different types of plants – for example, conifers and broadleaf trees and even between plant species.

NIR analysis was originally developed by the U.S. military during World War II to detect enemy camouflaged tanks.  It is now used for tasks such as crop and timber inventory, damage assessment after a forest fire, and to verify insurance claims after a hail storm.

The founders of GeoVisual hail from Microsoft, Google, and Tektronix. The company originally funded itself through a series of Small Business Innovation Research (SBIR) grants from NASA to analyze satellite data for global forest and agriculture monitoring. The company uses machine learning to improve the accuracy of its predictions. GeoVisual is now turning its attention to higher value commercial crops.  Orrey predicts that, as high resolution satellite imagery becomes more affordable and drone imaging gains widespread adoption, the fusion of these technologies with ground-based observations will enable the company to impact farmers’ profits.

Confirmation of that prediction came recently when Caterpillar, a company more typically associated with construction and mining equipment, diesel and natural gas engines, and industrial gas turbines, announced the release of the Cat S60 Smartphone. In addition to being ruggedized to be dust-proof, drop-proof, waterproof, and scratch-resistant, the S60 smartphone includes an integrated thermal camera from FLIR Systems. Although infrared thermography and NIR are at different places on the spectrum, the trend is clear.   Available later this year for $599, the S60 may eventually be used by farmers and agronomists mounted to a truck, essentially serving as a poor man’s satellite.  The potential market demand for analyzing data from such devices is huge.

According to the United Nations Food and Agriculture Organization (FAO), until the year 2000, agriculture, the world’s oldest business, was also the world’s largest business in terms of employment. Since then, the services sector has surpassed it in size. Although employment growth in agriculture has slowed, the number of workers in agriculture reached 1 billion in 2009.  Employing one out of every seven people in the world, subsistence farming remains the most common human experience on the planet. In the future, companies like GeoVisual Analytics may use AI to help these people and do well by doing good.

Comments are closed.