Sound Businesses: Profiles of companies and business models we are keeping an eye on.

Taranis combines military-grade optics with machine-learning imaging to provide farmers with millimeter-level insights into their fields, knowledge that can help increase crop yields and reduce food waste that amounts to $750 billion each year.

The Tel Aviv-based company’s precision agriculture platform is based on high-definition hyperspectral imaging from drones and aircraft giving a resolution of less than one millimeter – compared to about five meters for satellite imaging. It’s an early-warning system for incoming diseases, weeds, insects and fertilizer problems that cause about 30% of crop losses worldwide,” says CEO Ofir Schlam.

Globally, the United Nations Food and Agriculture Organization estimates that 1.3 billion tons of food – one third of all the world’s production – is lost or wasted each year, costing about $750 billion annually.

As much as 20% of some crops are lost even before they leave the farm Farmers have to battle 30,000 different weeds, 10,000 species of insects and an array of diseases caused by fungi, bacteria and viruses – all threatening the quality and volume of farmers’ yields, according to Bayer, a major producer of agricultural chemicals.

“What we bring the farmers is a real insight into their field. They already have great knowledge about farming and about their farm,” says Schlam. “We’re enhancing it.”

Growing Sector

Taranis is one of a growing cluster of Israeli agritech companies updating the country’s rich legacy of farming innovation that includes the invention of the kibbutz communal farm, drip irrigation and disease-resistant seeds. That forward-looking farming tradition is now merging with the country’s booming high-tech sector to leverage technologies first developed for military use and repurposing them for agriculture.

Global agritech investments in startups grew 44% to $6.9 billion in 2018, according to Agfunder. Investment in farm management software, sensing and IOT companies jumped 65% year-on-year to $945 million, with remote sensing startups dominating the sector.

Patented Technology

Schlam says that Taranis’s precision imaging is what helps set it apart from rivals. While Taranis does not develop its own cameras or drones, it invented the patented hardware that enables any high-end camera or sensing system to capture clear images while in fast motion, matching capabilities previously found only in multi-million-dollar systems in fighter jets.

“We can capture clear images down to the insect and leaf level from planes traveling at 125 mph and drones at 35 mph. Nobody else can offer that capability. The closest on the market is 1,400 times lower than our resolution,” he says.

Using artificial intelligence trained on a unique dataset of 10 million agricultural indicators, the platform can spot disease and pests before they are visible to the naked eye, and identify species of weeds before they are fully grown – warning farmers early enough to take the necessary action to prevent harm to crops.

Insights from the company’s software can increase yields by up to 7.5%, or $30 an acre, on crops like corn and soya beans where margins range between $30 and $50 an acre, Schlam says. “That’s quite a bit of value.”

Farmers pay Taranis a fee per acre equivalent to about 25% of the extra yield they can achieve with the system.“There’s a good return on investment for everybody involved,” he notes.

“This is huge,” says Yanai Oron, a general partner at Vertex Ventures which started investing in Taranis in 2017. “In this industry, a 7.5% change to the top line might mean the difference between losing and making money for the season.”

Taranis increased its total funding to $30 million in November by raising $20 million from a strategic partnership with Canadian agriculture giant Nutrien and other investors including Finistere Ventures, Viola Ventures, Vertex Ventures and OurCrowd. The deal was No. 8  on Agfunder’s list of the year’s top 20 farm management investments.

Taranis has deployed more than 100 of its imaging pods to third-party drone and aircraft operators in the U.S. – its biggest market – and Canada, Brazil, Argentina, Russia, Ukraine and Australia, monitoring a total of about 20 million acres of farmland.

The 10 crops that Taranis focuses on – corn, soya bean, cotton, wheat, potatoes, canola, barley, sunflower, sugar beet and sugar cane – represent about 70% of world farming, and 60% of global agriculture losses.

“Each pod can cover half a million acres a month. With more than 100 pods we can potentially cover tens of millions of acres,” Schlam says.

In addition to the Nutrien tie-up, in March Taranis announced an Australian partnership, combining its technology with the crop-protection solutions of ADAMA Australia and Landmark.

In 2018, the company acquired Mavrx, a San Francisco imaging company capable of taking high-resolution images with a fleet of 60 planes in 30 states across the U.S. that can recognize and classify weeds, diseases, insects and nutrient deficiencies on a single leaf at a resolution of 0.5 millimeters.

Taranis combines the imaging with a deep trove of data including proprietary weather forecast software, satellite images and other sensors to give farmers a broad picture of soil, climate, moisture and other information on which to base their decisions.

“What we bring the farmers is a real insight into their field. They already have great knowledge about farming and about their farm,” says Schlam. “We’re enhancing it.”

The information produced by the platform changes through the year. Early in the season, farmers get a first glimpse of stem count in the field so they can decide on nutrients or re-planting if necessary. Then the system tracks weeds, alerting farmers to the type of herbicides they might need and identifying resistant strains. As the crops grow, output will focus on insects, disease, more weeds and deficiencies in fertilizer and water.

With 19,000 customers in eight countries, the company is earning revenues but is not yet profitable. Investors are on board, says Schlam.

“Right now they see the benefit of growing the company further and investing the revenues back into the company and expanding our headcount,” he says. “The market is so huge we want to capture our early user advantage in different geographies. We have a few more years of growth when we intentionally don’t want to be profitable.”

“So far we’re seeing 3x growth year on year; we’re really happy about that,” he says. “We think we can sustain this growth rate.”

Beyond Diagnostics

The major obstacles to expansion are the conservatism of the agriculture industry and farmers’ small margins.

“Growing in agriculture with today’s commodity prices is definitely not easy. We are not the most important thing a farmer needs to grow on his farm. First they need feed and other input,” he says. Schlam wants to persuade farmers that it’s a worthwhile investment with an immediate return. “There’s a learning curve there because it’s a new type of solution. So time is basically the main risk, how long it takes to adopt,” he says.

Schlam sees potential to expand the company’s offerings based around its imaging, and more involvement in solving problems instead of just diagnosing them by connecting the platform to crop dusters and irrigation systems. “We keep seeing more options and better ways to leverage the technology,” he says. “Right now we are mostly focused on diagnostics for the problem.”

“We believe we can make it more actionable by integrating better with the machinery companies. That’s a great opportunity we are pursuing,” he says.

That would be the right direction, says Ewan McFarlane, head of digital agronomy at Origin Enterprises Plc. Much of the output from the 20 or so companies like Taranis offering remote-sensing agricultural solutions “is like looking in the rearview mirror. We can see something that’s happened in the past but it doesn’t enable us to make a management decision on the crop.”

“The earlier that you can make an observation – particularly about weeds, pests, diseases or plant population – and the higher the resolution, the more opportunity we’ve got of making a management decision which will impact the final outcome, whether in reducing costs or increasing yield,” says McFarlane.

Taranis and its competitors also must contend with the slow pace of farming, which is largely controlled by nature.

“The agricultural industry globally has always been relatively slow compared with other industries to adopt new technologies. A lot of investors will look at an exit in five to seven years but in agtech businesses our learning cycles are much slower because they are determined by the calendar rather than by how much money we push into it,” McFarlane says. “The cycles for data acquisition for big data mining and artificial intelligence are determined by the seasonality of our crops.”

Schlam says some farmers are using the platform successfully to assess new land they want to acquire. The company is confident enough in its technology to encourage doubters to try the system for themselves.

“We’re seeing clients that maybe were even a bit skeptical, who decided to implement the solution on say 10% of their farm, on a few fields,” he says. “After one season they are extending it to the full scope of the farm.”