A clear vision

Founders: Sudhanshu Rai and Sumit Sheoran
Year: 2019
Big idea: Developing an IoT device for precision farming

Computer science graduates Sudhanshu Rai and Sumit Sheoran were eager to establish a company that would make farming more profitable. They came from agricultural families — Rai was from Uttar Pradesh and Sheoran from Haryana — and had witnessed the many problems that farmers faced. While searching for viable solutions, they analysed nearly 8 million data units from the Kisan Call Centre, a platform that addresses farmer queries.

Rai, who graduated from the Dr. A.P.J. Abdul Kalam Technical University, Lucknow, and Sheoran, an alumnus of the Indian Institute of Technology Roorkee, found that almost 95% of the queries to the centre were about weather patterns, pest attacks, disease incidence, and fertiliser usage. They also observed that farmers engaged in horticulture asked more questions than those working on crops such as paddy and wheat – perhaps because horticulture required decisions to be made on a day-to-day basis. So, the two decided to focus on fruit and vegetable farmers.

They developed an initial version of an IoT (Internet of Things) device with sensors to collect real-time data on farm conditions, enabling them to provide advice to farmers. They started with five grape farmers in Nashik, but realised that the farmers, though enthused by the idea, did not follow it up because they did not think that two people in faraway Bengaluru would know more about their farm than they did. So, the two moved to Nashik to start work with local farmers. The move produced immediate results, with about ten farmers buying the IoT device and following their advice. Now, Fyllo — Greek for leaf — works with about 12,000 farmers in 11 States, with 18,000 installations. The bulk of the business is from Maharashtra, Gujarat, Karnataka, and Madhya Pradesh. They sell aboutt 3,000 devices a month. The farmers buy the IoT device (priced ₹5,999 and above) and pay an annual subscription fee (of about ₹2,000 a year) for the software.

The start-up has developed an upgraded version of the IoT device, named Nero Infinity – Nero is the Greek word for water. It measures soil and weather parameters and advises farmers on issues such as soil moisture status, when to irrigate, and fertiliser use on a mobile app in local languages. Fyllo's 150-strong team includes agronomists, crop physiologists, data collectors and analysts, software professionals, hardware engineers, and sales and marketing personnel. They regularly collect ground data to train their models. 

The two-and-a-half-foot-long Nero Infinity device has to be buried in the ground for soil data. Each IoT device can cover up to 1 hectare, Sheoran says. Another product, called Kairo, which in Greek means climate, provides data on climatic conditions.

Founders: Sekar Udayamurthy, Krishna Iyengar and Vinodh Venkatesan
Year: 2018
Big idea: Using computer vision and artificial intelligence for quality assurance in manufacturing

Krishna Iyengar faced a problem when he headed manufacturing in an auto component company. It troubled him when clients complained of quality issues in a finished product. "I was at the receiving end of customer complaints," he says. 

Iyengar, who holds a dual degree in Metallurgy from the Indian Institute of Technology Madras and a PhD in Theoretical and Applied Mechanics from Cornell University, examined methods to detect and eliminate defects on the manufacturing line itself. The quest prompted him to team up with Sekar Udayamurthy, who has a Master's in Information Systems from the University of Maryland Baltimore County and has extensive experience in the information technology industry; and Mechanical Engineer Vinodh Venkatesan, who was Iyengar's colleague at the auto components company.

The three started Jidoka Technologies to automate visual inspection and quality control in manufacturing. But while quality control slowed production, faster operations resulted in more defects. The problems included high rejection rates, manual sampling inspections that slowed down manufacturing and still did not catch all defects, and outdated vision systems. Iyengar says Jidoka has developed an artificial intelligence (AI)-powered computer vision that offers accuracy rates as high as 99-99.5% with 100% inspection. The company has two products that provide cognitive product inspection and intelligent process optimisation.

One of the products, Kompass, is an AI-powered vision system that can inspect 10,000 images in a minute and help in instant decisions on rejecting defective parts. "We specialise in instantaneous decisions," he says. The system takes an image, applies AI interference, and sends a signal to reject a faulty part in the production process. "All this happens within a space of 100-150 milliseconds," he says. Jidoka utilises German cameras to capture visual images and develops its algorithms in-house. It provides a turnkey solution to its customers, which include a leading car manufacturer, a global alcoholic beverages company, and a leading biscuit-maker.

Another product, Nagare, focuses on process inspection, which leads to improvements in production system and line efficiency. He cites the example of kitting in an automobile manufacturer, where different parts required on the assembly line are put together into a box in the warehouse. Each part is numbered as per the production sequence, and even if one part is missed out during kitting, the production is affected. Jidoka utilises visual inspection, while protecting the privacy of the worker involved, and ensures that such mistakes do not occur, he says. Kompass hit the market in 2019, and Nagare in mid-2023.