BANGALORE -- Krishna Kumar knew he wanted to do something more meaningful with his career. So nine years ago the programmer quit his job at engineering giant GE intending to help improve the livelihoods of farmers in his native India. When he started looking at the agriculture sector, he found that despite its huge scale and economic significance, it was deeply fragmented and poorly documented.
"The whole ecosystem is far-flung. It is very difficult to get to the root of any problem and solve it," he said. "The biggest gap was the data."
In 2010, the Bangalore-born entrepreneur launched CropIn Technologies in his home city, providing a platform to link smallholders with buyers, gathering information about farmers' yields, practices and ground-level challenges. As the platform grew, so too did the amount of data that the company had on Indian agriculture.
By 2016, CropIn's data set covered 2 million hectares and more than 250 different crop types. Using machine learning, CropIn analyses how farms performed under various conditions, from pest outbreaks, to disease, to water scarcity. That data is turned into predictive models to identify problems before they occur and provide buyers with accurate information about future yields.
"Now, three months before the harvests, [buyers and traders] know how the farms are behaving and what they can promise to their customers," Kumar said. "At the same time, if you know that yields are not going to be high... you send your agronomist to the farms and help them get the right yields."
CropIn is one of a growing number of tech companies riding a wave of interest and investment in 'digital' or 'precision' agriculture-data-enabled farming that uses sensors, drones and the internet of things to optimize inputs and track commodities from farm to fork.
The world market for precision agriculture is forecast to hit $12 billion by 2025, according to research company Global Market Insights, and many of the industry's largest participants are moving into the space.
In 2017, fertilizer company DowDuPont spent $300 million on Granular, a developer of farm software, while tractor-maker Deere & Co. paid $305 million for agricultural robotics company Blue River Technology. Back in 2013, agriculture giant Monsanto paid nearly $1 billion for weather data business Climate Corp. Others are developing their own in-house systems. Bayer launched its own digital farming business, Xarvio, in late 2017; Yara followed suit a few months later, launching Atfarm in March 2018.
This push into data is partly motivated by the agri-food industry's need to better understand and manage the long and complex supply chains that link farms to consumers.
Worldwide, consumers' trust in food standards has been damaged by scandals in recent years. In 2013 ready meals supplied to U.K. supermarkets were found to have been adulterated with horse meat. In the U.S. and Canada huge quantities of romaine lettuce had to be destroyed last year after some was found to be contaminated with e.coli bacteria. In China, an outbreak of African swine fever has led to more than a million pigs being culled, and sparked worries among Chinese consumers over the safety of their meat.
Environmental and social concerns are also starting to have an impact on food companies' bottom lines. Southeast Asia's palm oil industry is under pressure to tackle deforestation, which has sparked consumer campaigns in Europe and a diplomatic spat between the EU and Malaysia, one of the crop's largest producers.
But tracing accountability in the complex global food system is difficult, when small, remote producers sell crops to wholesalers and ingredients have to cross continents multiple times on the way to a supermarket shelf.
AI and big data could change that.
"There has to be trust built, and the best way to build the trust is to connect the dots, show how it was grown, where it was grown and its authenticity," said Claire Pribula, head of Asia-Pacific for the Yield Lab, an agricultural technology incubator which opened a program in Singapore last year. The Yield Lab's U.S. and European arms have invested in data and sensor technology companies, from real-time monitoring of beehives to blockchain-based traceability systems for horticulture. "There's such an integration of technology happening. It's amazing, and it's all moving in the right direction."
Pierre Courtemanche, CEO of Geotraceability, an early pioneer of farm-level traceability, which started out using machine-readable paper surveys in Ghana, has recently begun working with palm oil buyers in Southeast Asia. He says that, although radically improved transparency in supply chains can be worrying for companies who have previously hidden behind the complexity of commodity markets, it is becoming a competitive advantage.
"When you're facing a shelf in the grocery store. [The products] are competitive, the price is about the same. So you think: 'Which one should I take?' And this is where traceability should help you in your decision. this is the big game changer," he said. The technology is close to the point where the buyer of a cup of coffee in a mainstream outlet could have a direct link back to the individual farmer who grew the beans, he said.
At CropIn, Kumar agrees that clients are coming to see traceability as a competitive advantage. Since its launch CropIn has picked up contracts in 29 countries. Investors too are taking an interest. In November 2018, CropIn raised an $8 million Series B round from Chiratae Ventures and the Bill and Melinda Gates Foundation Strategic Investment Fund.
"Most people think of this is a cost center, but I think it will become a profit center because traceability will not only give you the ability to make your consumer happy, but it will also help you to look within your supply chain," he said. "If you are in the business of dealing with farm produce and you are not in the business of traceability, you are not protecting your brand. You don't know what promises you are making to the market."