Deep Learning in Agriculture

Spandan Samiran
Jun 15, 2018 · 4 min read

“I believe in the future of agriculture, with a faith born not of words but of deeds.” This is a famous quote by E.M. Tiffany. ‘Our deeds’ will shape the future of agriculture. Coming of age technologies like AI is turning the future of agriculture in ways we could never envisage. Splurge in population, the ever-changing climate and deteriorating ecosystems along with the pressure of food security and sustainability, and harvesting at least 4 times of what we are harvesting at the same time.

Addressing these problems? Difficult!

But not impossible, at least not anymore.

Fascinating how an intricate and complex agricultural ecosystem is now being monitored, analyzed and measured to understand its physical aspects and phenomena. There are several technologies that have carved their way into agriculture. But the one earning a lot of attention these days is “DEEP LEARNING”, so what is this deep learning and why is it gaining momentum?

Deep learning is a class of machine learning algorithms. A deep learning technology is based on artificial neural networks (ANNs). As a part of artificial intelligence (AI), deep learning stands behind many innovations. DL also known as hierarchical learning provides “deeper” neural networks that provide a hierarchical representation of data.

Now why DL is the next big thing and why should this be incorporated in agriculture?

Agriculture is a particularly complex area of application. Every area has its own features and climate, even that change with time. So, the farmers look for that one best trait. They search for features of significance like a better use of water and nutrients, adoption to the change in climate, and resistance to disease in a growing crop. That takes up a lot of data to analyze, keeping in mind the real-time changes. This makes Deep learning the revolutionizing breakthrough that the agriculture needs right now. With the aid of these algorithms, plant breeding is becoming more accurate, efficient. When the algorithm is fed with field data and insights about how a variety may perform when faced with different sub-climates, soil types, weather patterns, and other factors; it builds up a probability model. When tracking any disease, early and accurate identification is essential. After sorting through a decade of images of diseased plants, this algorithm can spot disease type, severity. Same goes for the weather pattern. All these are digital testing and they can’t replace the field trials but can provide us with a predictive vision. The core aim of modern agriculture is to create seeds and crop protection products that provide relief to the posed global challenges. And switching to DL is coming one step forward towards a better Global market.

Why not traditional machine learning? Why switching to DL?

In traditional machine learning, the learning process is supervised, and the data needs to be labelled before feeding to the computer, this is where DL can be beneficial. The main advantage of DL is that the program builds the feature set by itself without supervision. Unsupervised Learning makes us more equipped and well-informed to work in a real-time environment that is unpredictable and ever-changing. This is important as the internet of things (IoT) continues to become more endemic, because most of the data humans and machines create is unstructured and is not labeled.

Once into DL, what are the points to keep in mind?

Well, there are some points to be pondered on when implementing DL in agriculture such as data variations, data pre-processing, data augmentation and the architectures we are using which are for depth study.

Surveys also indicate DL out-run most of its contemporaries when compared under same datasets, which are also for further depth study into DL.

A real cut to edge technology. However, this will remain a thought if we don’t work upon it. Problems? Will come, Limitations? Will be there, but that should not keep us from diving into something that can change our future, that can evolve the ways of agriculture. Our future is calling, the question is how do we respond, how could “our deeds” carve out a new tomorrow, where we will be ready for whatever comes. Let’s preserve what needs to be preserved and let’s perfect what can be perfected. We’ve always heard “Change is the only Constant”, so if next time the weather or type of disease or any other aspect change you can just say that relax because rest is taken care of.

Join us in the search for a better tomorrow, we are here to offer advancements that have potential to change the way we eye agriculture.

Spandan Samiran

Written by

Founder of FarmatroniX — aims to automate agriculture so that we can make every person food sufficient