Manohar Sambandam has spent nearly 27 years in the semi-conductor industry working to tight deadlines. But, now as the CEO and co-founder of Green Robot Machinery Pvt Ltd (GRoboMac), he is racing against time to accomplish something by July. He also has to convince investors that his project – building a robot that will pick cotton – will indeed be capable of picking as much as cotton as it can, as he seeks to raise over ₹2 crore.

In a small office that he shares with another company in Bengaluru’s Whitefield, Manohar demonstrates the movement of the robot’s single arm. That arm should pick 15-20 cotton buds a minute and with multiple arms, by July, Manohar will have to show that it can pick 250 kg of cotton a day. That is the challenge he is up against. “The need is there and the technology is available. I have to prove that it can pick 250 kg of cotton a day,” he says.

After graduating in Physics from the Madras Christian College, Manohar completed his bachelor’s and master’s in electronics communication from IISc, Bengaluru. He was in the semi-conductor industry, including a 14-year stint at Texas Instruments.

Passionate about farm tech

When he turned 50, he wanted to do something on his own and he looked around for ideas. He had bought some agricultural land near Tiruvarur in Tamil Nadu, as he was passionate about agriculture and believed technology and proper farming can increase farm yields. “I wanted to prove that it is possible to earn up to ₹1 lakh an acre,” he says. He soon realised what he was up against, a shortage of farm labour being a major problem.

“Labour is not available. It is not about the cost of labour. I was looking for a problem to solve. Why don’t I solve this problem of picking,” he says. Across the country, the issue was the same. He chose to try and solve the problem of labour shortage for cotton as he believed it was a crop that would help him realise his objective of proving that it is possible to earn ₹1 lakh an acre.

He quit his job in 2013 and started the venture in 2014. “I looked at it purely from a technology perspective. Can I solve this? Can I replicate the human way of picking?” Manohar looked at machine vision, basically stereo image, to find out if it can detect and locate the cotton pods that are ready for picking. He was convinced that technology was available and that it can be adapted for what he had in mind and that there is enough computing power available to achieve what he wanted to. Reliability and robustness were a challenge.

With help from professors at IISc, he set about building a robot to pick cotton. By August 2015, Manohar had a proof of concept ready, to detect and locate the cotton. He and a colleague from Texas Instruments put in about ₹6 lakh to kick off the venture. Manohar used interns from the engineering colleges in Bengaluru to work on the project, even hiring a few of them. Once he got the proof of concept ready, he believed that investors would find his venture attractive to put in money and he tried raising funds, but found that was not possible until he had a product ready. He got a ₹35-lakh grant from social enterprises incubator Villgro.

Farming-as-a-service

GRoboMac’s robot will have a seven degree arm. Manohar is now working to have a product ready that can prove it can pick 250 kg of cotton a day. He has had discussions with farm equipment manufacturers. He is targeting a 2018 launch of the robot, one that can pick 500 kg of cotton a day, which is the equivalent of 10 labourers. He expects it to cost ₹3.5-5 lakh and will be bought by tractor manufacturers and others who are into providing farming-as-a-service.

The money he hopes to raise will help him recruit engineers with qualifications and expertise in machine vision and robotics. The robot will have a huge amount of embedded software. Manohar also believes that the robot can be adapted for any horticulture crop. “The business model was a challenge. Now, with companies in the farming-as-a-service (FaaS) space, it is getting organised. The need for something like this is high. Agriculture is not able to scale because of labour issues,” he says.