Understory’s Move into Agriculture Comes into Focus with Monsanto Partnership

After one year of testing, Monsanto is collaborating with hyper-local weather data provider Understory to optimize the company’s seed production operation in Argentina.

In early 2016, the weather sensor hardware and hyper-local data network company raised $7.5 million in Series A funding in a round led by Monsanto Growth Ventures and 4490 Ventures, an early-stage venture fund based in Wisconsin. 

Understory detects weather events at the earth’s surface instead of in the atmosphere like traditional weather stations, to create hyper-local weather reports and the investment from Monsanto marked the startup’s first move into agriculture. It was unclear where this investment would lead until now.

Understory and Monsanto recently completed a year-long trial of Understory’s technology in Hawaii to improve the yields of Monsanto’s seed production operations. Though not a bastion of broad-acre agriculture, Hawaii’s climate allows for unique research opportunities.

“The temperature is perfect for corn and soybean testing and you’re able to have complete control over when you plant and when you harvest. It really allows for the ability to put three growing seasons in one year,” said Understory CEO Alex Kubicek.

Understory’s core business has been providing damage analytics to P&C insurance industry to improve the claims experience for home and commercial properties. In the United States Understory networks cover five major metropolitan areas, including Boston, Dallas and Kansas City, with plans to expand to 5,000 stations by the end of 2019. The patented technology provides actionable information that radar, satellite and other remote sensing technologies fail to deliver. Understory is looking to expand further into agriculture but has an exclusive deal wth Monsanto in Argentina.

“Understory provides the only technology necessary to make hyper-local weather valuable for operations management in agricultural areas where local weather information is not accurate,” said Giovanni Piccinni, global supply chain field optimization lead for Monsanto. 

Each station in the network provides 50,000 measurements a second to power an artificial intelligence core that provides precise hail, rainfall, temperature, humidity, evapotranspiration (moisture leaving the plant) and growing degree units (where a plant is in its lifecycle to determine best harvesting time). Understory data informs Monsanto’s decisions around irrigation, harvest dates and times, and seed production.

Piccini said that the major hurdle of success throughout the year of trialing Understory was predicting rainfall.

“I’ve been working with weather stations for a long time and the biggest hurdle is always being about how to estimate accurate rainfall,” explained Piccini, and this is where Understory succeeded in trials. 

“We’re in the middle of really understanding how our data can allow a supply chain to make decisions ahead of time about where seed is going to be produced. We believe that will be of incredibly high value to the partners we’re working with,” said Kubicek.

Understory’s major differentiator is that it’s weather stations have no moving parts, meaning they require less maintenance – a major attractor for Monsanto.

“We are not in the business of managing weather stations,” said Piccini. 

Understory sites, installs, and operates the network, while providing a real-time data platform and API. In addition, Understory’s weather network can be integrated with any third-party sensors currently available on the market.

“Understory at its core is an infrastructure company where we own and operate weather networks and apply them to improve decision making.” Kubicek. 

A Growing Set

Weather-station based startups are a growing subset of agtech with various outfits operational in India, the US, and the UK. Piccini explained that competition for coverage in untouched markets is getting fierce since the most opportunity is likely to be found in areas without effective publicly-funded weather data. 

“Weather information is key. There has been a mad rush everywhere. South America and Europe are where there is the biggest lack of accurate weather,” said Piccini.

Kubicek explained that though weather station startups are popping up all over the globe, the various companies could end up partnering instead of competing, as one proves to have superior hardware and another, enhanced analytics.

Piccini added that though remote sensing capabilities in this field are improving, the level of granularity needed to direct a large seed production operation like Monsanto’s is too expensive at this point since aerial flight-collected data is the only acceptable type at this point and Monsanto’s operations are too vast for flights to be cost-effective.

Startups in the local weather station space include:

Skymet Weather Services has a network of 6,000 automatic weather stations (AWS) in India providing farmers with hyperlocal forecasts to help with decision-making. The company also serves as a fintech platform in rural India for insurance companies and banks, providing claim settlement and product design.

UK-based startup KisanHub uses big data analytics, cloud computing, and machine learning to compile data from satellite imagery, weather stations, soil sensors, and other sources. The startup is planning to use its latest round of funding to deploy 100 of its own weather stations across the UK.

In 2016, precision ag data services provider Farmers Edge launched a collaboration with IBM’s The Weather Company. The partnership allows Farmers Edge to integrate hyper-local forecasts from The Weather Company’s Forecasts on Demand (FoD) weather forecasting engine into its predictive modeling software. 

In a similar deal, satellite imagery company Geosys, partnered with German weather station company Pessl Instruments to provide clients with more localized weather data. The global partnership incorporates data from Pessl Instruments’ FieldClimate platform into Geosys products, such as its monitoring platform Croptical.