On a sunny November morning, Ananth Erasappa and his team were setting up their drones in the middle of a corn farm. They had just travelled to this cluster of corn farm in Gogulampadu in Andhra Pradesh’s Krishna district.

The day would be long for Erasappa, the chief flight operator at Bengaluru-based Skylark Drones. He and his team had to survey nearly 150 to 200 acres of corn farms by the end of the day along with Swiss agribusiness company Syngenta’s India unit. On the farms stood crops from Syngenta’s seeds.

“The most challenging part was to identify the farms and cover them… most of these farms are distributed across villages with each farm size at an average of anywhere between one to two acres,” says Erasappa, who spent well over 20 days conducting the exercise in villages across the region.

According to a FICCI-PWC report, around 15 million farmers in India are engaged in maize or corn cultivations. Corn is also considered to be the third most important food crop for India after rice and wheat. The aim of the Syngenta-Skylark project was simple: count the number of germinated seedlings and plants across corn farms using Syngenta seeds with an intent to see how to increase productivity.

This is not the first time Skylark Drones and Syngenta are collaborating. They had started working together in 2017. “Last year in December, we were parallelly conducting the crop counting exercise in the Philippines for Syngenta while the Indian project was happening,” says Mughilan Ramasamy, co-founder and CEO of Skylark Drones. “This year, we plan to do the project around October. We plan to change the operations model a little bit. We plan to send both the teams on site itself for data acquisition as well as analysis.”

Counting the cobs

One of the first challenges was to group farms together based on proximity to one another. “We had a target of around 150 to 200 acres to be covered in a day.. we had to first figure out the layout of the land parcels. It was not like 1,000 or 2,000 acres of land in one patch,” says Ramasamy. The dispersed nature of land holdings didn’t help, in short.

To start with data acquisition, the team initially conducted a sample survey using drones “We initially conducted sample surveys in three to four farms using drones and then cross-checked the results manually. We were able to achieve 90% to 95% accuracy in these sample tests,” says Erasappa.

A typical day at the project would begin with identifying the farm’s boundaries from a drone flying high. A coordinator from Syngenta who worked with the drone team helped identify the exact boundaries of the farm on an iPad.

How? Check the colour of the farm. Plants grown of Syngenta’s corn seeds tend to have a particular shade of green which is different from the neighbouring corn farm of different seeds.

This lends itself well to detection of the crop health based on images. For example, the colour – or, more accurately, the wavelength of light – reflected by leaves of a plant at different stages of growth or health vary. Drone- captured images, which captured reflected light, then, can be used to analyse the health of a crop.

Back to the farm’s boundaries. Think of the images captured here as the multiple images stitched together in a panoramic image on your phone camera. The stitching can be properly done only if the edge portions of each image overlaps with the adjacent images – giving the software a proper reference area to stitch together. The lack of a proper overlap will lead to patchy or non aligned image which will corrupt the final image: whether it is a panoramic photo on the phone or an image in the crop counting project.

After this, the data from the drones were downloaded and verified on an app created by Skylark for Syngenta and, once authenticated, the data was shared in real-time with the drone company’s Bengaluru team for processing and data analysis.

Drone doctors at farms

While Skylark Drones and Syngenta’s project looked at the crop counting another drone project across the border in Telangana was playing doctor to farmers.

Hyderabad-based engineering design services company Cyient and a leading fertiliser company in south India are working together on a precision agriculture project to monitor farms using drones in order to monitor and act on crop health.

“Drones are being used for data acquisition at the farms and can also be used for delivery of certain fertilisers, pesticides etc..,” says Dinakar Devireddy, head of innovation program at Cyient, asking for the fertiliser company client to be kept anonymous in this story.

The proof-of-concept (POC) was conducted at Mahabubnagar at an area covering seven to 10 acres consisting of different crops a year ago. After the POC, the team was able to see an improvement in the yield that prompted it to scale up – the project to begin in Nalgonda district in Telangana in a few weeks.

“At the end of the PoC project, we were able to see a 20% rise in the yield. Now, in the next phase we have increased the area to 2,000 acres,” says Devireddy.

The idea is to identify areas that are stressed, either due to a nutrient deficiency or affected by weeds or pests, and alert farmers in the form of digital prescriptions.

Hurdles before drones in agri-tech take off

These might just be a few examples of drones being used in Indian agriculture but the bigger question is whether this can scale up.

One of the biggest hurdles to cross is the regulatory challenge before drones in India which has been left in a state of chaos with delays coming in one after the other. More than two years have passed since India released its first set of guidelines for use of drones and unmanned aerial systems. The final regulations were said to be released by early 2018 but a to-and-fro between multiple ministries has led to delays in its finalisation and implementation — causing uncertainty and confusion in an industry expected to generate millions of dollars in revenues and thousands of jobs.

Besides regulatory hurdles, there is also the challenge of using drones for spraying fertilisers or weedicides, which require a certain level of skills, says Abhijit Joshi, Senior Vice President at Jain Irrigation Systems. Still, he feels that drones, while new in India, offer a big scope to improve agriculture outcomes given that they can cover large areas.

Another challenge Joshi points out is the fragmented nature of farms in India and the affordability of a drone-based service. “In India, field sizes are fragmented. Where one farmer cannot afford the drone service, if the government or community comes up with the scheme it can become viable,” he adds. Indian agriculture is replete with examples of such shared infrastructure whether it be heavy tractors, threshers or other machinery.

Even though there are challenges here, globally there are many use cases of drones being used in agriculture and there are many companies developing products for the same.

Joshi talks about an example where a drone is programmed to routinely take off to scout or survey a farm on a pre-planned route to check and monitor crops after which it returns to the docking station to download the data captured. With the speed and paths across which a drone can fly, it surely makes for a quicker and easier way to monitor large areas compared to a manual scout.

---