Describe the descriptive analytics applications of TellusLabs.
Read the case study and answer the following questions. Bullet points are fine.
TellusLabs
In 2016, the United States produced more than 15 billion bushels of corn and 4 billion bushels of soy, making corn and soy two of the most widely produced crops in the country. These large yields affect more than 3 million U.S. farmers, as well as thousands of decision makers in finance, insurance, agribusiness, and government. With so much at stake, accurately predicting crop yields has become critical.
TellusLabs (www.indigoag.com/geospatial-innovation) is a Boston, Massachusetts, start-up recently acquired by Indigo AG that integrates data from three sources for analysis purposes: daily satellite imagery from the National Aeronautics and Space Administration (NASA; www.nasa.gov) and commercial satellites; weather data from the National Oceanic and Atmospheric Administration (www.noaa.gov); and seasonal, crop-growing information from the U.S. Department of Agriculture (USDA; www.usda.gov). The firm uses machine-learning algorithms (see Technology Guide 4) in its analytics efforts to generate intelligence about natural resources; for example, predicting agricultural yields.
There are other satellite imagery analysis firms such as Descartes Labs (www.descarteslabs.com) and Orbital Insight (www.orbitalinsight.com). However, TellusLabs differentiates itself by applying scientific expertise in vegetation and climatology to its analyses and by focusing exclusively on natural resources. The firm maintains that there are many organizations that must make difficult decisions about natural resources. Its objective is to quickly provide quality data to them.
The company’s first product is Kernel, an agricultural commodities forecast modeling tool. Kernel provides an interactive, online dashboard that shows a map of the major corn- and soy-growing regions across the United States. The dashboard also displays key financial indicators, such as predicted yield, harvested area, and total production. Users can view data at a state, an agricultural district, or a county level. In addition, they can examine historical yield data from the USDA. The dashboard also indicates the average change in corn and soy yield estimates, week over week. TellusLabs updates these forecasts daily.
Kernel is designed to provide fast, reliable data that organizations and people can use in multiple ways. For example, a commodities trader could use the information for trades in the futures market. An ethanol plant operator could consult Kernel to gauge whether its contracted farmers will be able to supply enough corn to keep it running. An agribusiness company such as John Deere could license the data feed from Kernel and integrate it into a smart pump that automatically adjusts how much water it provides to crops.
TellusLabs asserts that an internal test revealed that the firm is able to predict the end-of-year yield for U.S. corn and soy more accurately than the federal government. In the test, the lab used its algorithms to analyze publicly available, historical USDA corn and soy yield data from 2004 to 2014 in order to predict year-end figures. Over that period, the lab’s estimates outperformed those of the USDA almost 70 percent of the time during the months of August and September, which are the key trading months for corn and soy. In 2016, Kernel predicted the USDA’s final 2016 corn and soy yield report two months ahead of publicly available in-season forecasts. Kernel exactly predicted the USDA estimate for soy, and it came within 1 percent of the USDA estimate for corn.
TellusLabs plans to release a predictive model for wheat; to extend its corn yield data to Argentina, Brazil, and China; and to monitor forests and large, fresh-water reservoirs by satellite. The founders note that there are many interesting, global-scale, geospatial questions that have not been analyzed. For example: How will we feed a rapidly growing global population? How will we manage our increasingly limited water resources? How can we best manage our land and forest resources in an era of global warming?
On a contrary note, the U.S. Department of Agriculture said in 2018 that satellite images by themselves still could not be relied upon to predict annual corn, wheat, or soybean harvests. Instead, the government’s main source of information remained farmer surveys and random field samples.
And the bottom line? In January 2017, the lab announced that it had raised $3.1 million in seed funding. Further, by September 2018, some 1,000 organizations and individuals around the world were relying on TellusLabs’s data. In late 2018, Indigo Ag (www.indigoag.com), a company helping farmers sustainably feed the planet, acquired TellusLabs. Today Indigo AG continues to grow this technology.
Questions
- Describe the descriptive analytics applications of TellusLabs.
- Describe the predictive analytics applications of TellusLabs.
- Describe a possible prescriptive analytics application for TellusLabs.
