We are SoilHackers :)
Developing robots to make farming more profitable and sustainable.
In 2015, I completed my Bachelor’s degree in Computer Science. During my graduation I fell in love with Programming and it fascinates me that how with programming we can make computers do anything we want.
After my graduation I started working as a Freelance Software developer and during my journey as a freelance I learned a lot of things other than programming like how to pitch proposals to clients understanding their requirements and take their ideas and develop softwares.
I come from a family whose primary occupation is Agriculture so since from my childhood I am very connected to farming and experienced the entire process of growing , harvesting food season after season. And I am very much aware of the challenges that farmers face in their day-to-day life.
So during my freelance career I came across this Machine learning thing and in IT field everyone is talking about Artificial Intelligence , self driving cars and all that. It fascinates me and I started learning more about Machine Learning.
In Machine Learning I find Image processing very interesting, like how we can train machine learning algorithms which can interpret the context from images and we have all the inexpensive technology available so anyone can use these algorithms on low end devices.
In that moment it clicked into my mind if Machine learning is changing every field like Healthcare, Transportation, E-commerce, and then why not Agriculture?
At the same time, I am aware that the ways we are practising farming are very much messed up and badly impacting our environment and degrading the health of soil.
And answer to these problems is Organic Farming but we rarely see any organic farms at large scale so I started meeting some organic farmers and try to understand the challenges they face while growing organic produce.
So many organic farmers told me they managed to make organic fertilisers , organic insecticides, but they can’t make organic weed killers and for weeding they have to rely on manual labour and that adds an enormous cost to their production.
At that time it came into my mind that we can train Machine Learning algorithms to recognise crop plants and do weeding with the same precision as manual labourer do, and such a solution can greatly reduce the cost of production for a farmer.
Weeding problem is not only for organic farmers even the farmers who are spraying chemical weed killer on their farms are facing challenges to control weed, as most of the weed species became resistant to weed killers and farmers have to spray more and more chemicals which effects the crop health and increases the production cost.
In 2017 I discussed the problem with a few of my friends, Daman and Mohit are both from the computer Science background and I know them since my university days. We started out with different prototypes and eventually settled on building Robotics solution because we are already going to use computer vision so it’s worthy if we can make an autonomous electric machine which will be also good in reducing carbon emissions as most of the existing agricultural machinery is fuel dependent.
Finally, in 2020 we finished our first working prototype and did some trials and got very positive results. Now we are scaling the solution to give its benefit to a large number of farmers.
- Charan (Founder, SoilHackers)
Amritsar, Punjab, India