Indian startup to use artificial intelligence to deliver hyperlocal info on farming

A startup founded by two Indo-Canadians will start field trials for their product this month, and in their case, field is the operative word since the new company plans on taking artificial intelligence to the farm.

Ukko Agro, named after the Finnish God of agriculture was founded late last year by Ketan Kaushish, who is from New Delhi, and Avi Bhargava, from Kurukshetra in Haryana. It uses the tagline, Helping farmers grow more, sustainably. Its cloud-based predictive system will be deployed in a limited commercial rollout with select growers in three provinces of Canada and five US states.

“What we plan to do it is helping farmers optimise crop inputs, which is pesticides, water, and more. We collect hyperlocal weather data, both observed and forecasted. We analyse this via machine-learning based algorithms. Then we tell farmers up to three days in advance when to spray and how much to spray,” Kaushish said.

The focus now is on pesticide usage, informing clients when to spray and the dosage that will most effective. “The primary goal is to help growers minimise crop losses so it’s a risk averse system,” Kaushish said. The next phase will involve taking the model to irrigation, telling farmers when, how much, and how often to water their crop.

Rather than being an app, farmers will be able to access information via periodic text messages, Kaushish said. “We know that farmers as not tech savvy, so our goal is to deliver value in a very simple way so that we don’t take away time from a farmer’s busy schedule. Our philosophy is that a solution should be simple enough to operate and advanced enough to facilitate on-farm decision making with minimal time investment from farmers,” he explained.

After North America, Ukko is targeting the European market. India is a fertile territory as well, and certainly on its horizon.

This is important, Kaushish said, with regard to irrigation: “That is why it is even more important in India because a lot of farmers don’t have irrigation equipment, they need to lease it or rent it. If farmers knew when it’s expected to rain, how much it’s expected to rain, how much moisture the soil’s holding as of now, they would be able to make more intelligent decisions.”

But India poses a complex challenge for a farm-specific AI solution like Ukko. The area of a single farmer’s holding in North America could be 15,000 acre, but that could mean 15,000 different plots in India. That’s one reason Ukko is looking at partnerships before entering the Indian market, including those with government departments.

Ukko Agro’s base model was developed by William Fry and Ian Small at Cornell University in upstate New York. Its core Toronto-based team, other than Kaushish and Bhargava, include engineers Tasos Stamadianos and Haardik Haardik.