Amazon Web Services (AWS) is targeting agriculture data interoperability and value creation by partnering with agriculture infrastructure provider Leaf for its unified farm data API. With the availability of Leaf's machine data translation capabilities on the cloud hyperscaler's marketplace, "customers across the industry can immediately deploy [the API] within their own environments and make their application compatible with over 50 different types of [agricultural] machine data," AWS Head of Agriculture Elizabeth Fastiggi told SDxCentral.

"It's that ease of use and being able to streamline procurement of software and other capabilities, but really being able to easily deploy that in their own environments and making it available across the industry," she explained.

According to Leaf CEO Bailey Stockdale, the company was driven to bolster its relationship with the cloud provider and reside in its marketplace "because of those very deep relationships that AWS has with OEM machinery providers in the market," Stockdale told SDxCentral. The partnership also "adds a level of assurance" for customers because Leaf had to pass "months worth" of security and architecture reviews as part of this process. Stockdale noted that's "ultimately a really positive signal for these companies who already are heavily invested in AWS architecture."

While the API is currently just available on AWS' marketplace, and the cloud provider "has been fantastic as a partner," Stockdale said partnering with other major cloud providers isn't out of the question. "We're delivering, I would say, critical technology to help push the industry forward and we want to make that as broadly available as possible," he said. "We're really excited for pushing this out as far as we can possibly go with [AWS'] user base, but our goal is to enable everyone. We want to drive the industry forward as fast as possible," Stockdale said.

The data-related challenges facing the agriculture industry surpass those faced by industries like finance and banking, Stockdale argued. Data from agricultural machinery, weather stations, and soil sensors, for example, tends to be in a proprietary format and difficult to read or combine.

Agricultural data also has a density problem. Geospatial data from a concentrated animal feeding operation (CAFO), for example, might be five gigabytes, and a farm field might produce terabytes of data, according to the CEO. Agriculture OEMs, for example, have "massive" and "extremely complicated" AWS cloud accounts due to the "astounding" amount of data in action, Stockdale noted. "You have to be super efficient in how you process that to extract any value from it," he said.

Another "specific and unique" roadblock associated with agricultural data is that most of it is unidentifiable, he added. "If you have a weather map and a harvest map for a field, there's no social security number. There's no unified ID that those link to. You have to do that association geospatially yourself," he said – a task that requires significant amounts of processing, compute, and storage to "derive any meaningful value from it."

Stockdale sees a trillion-dollar market opportunity for agriculture insurance and lending based on the "massive opportunity for first digitizing a lot of this data and then being able to use the data for new use cases, like bacteria or carbon markets or just modernizing [and] validating claims," he said. "It's really an untouched industry and has massive growth potential."

AWS last year launched an agriculture solutions library with responses to more than 26 use cases in seven different areas. "We have solutions there within agriculture science, within precision agriculture, equipment manufacturing, grain and protein processing, supply chain sustainability – these are all really important areas for the industry," Fastiggi noted.