Modern agriculture has come a long way over the last two decades. With many technological advancements, the practices in farming have evolved from traditional methods to digital tools. Now, advancements in machine learning and artificial intelligence is being used in this field to ensure growing demands are met by optimising resources.

The Indian government think tank NITI Aayog had recently unveiled a discussion paper which addressed the national strategy on AI and other emerging technologies to be focussed on five core sectors. Agriculture was one of the key sectors mentioned in the draft, because the use of new tech would enhance farmers’ income, increase farm productivity and reduce wastage.

In this article we are going to discuss ML and sensors which can bring out the best from farming practices.

Agro Data Drives ML

In the recent MIT Technology Review’s EmTech Conference, Sam Eathington of The Climate Corporation, tells how sensors will become ubiquitous in farming in the coming days. “In the next five to 10 years, we’re going to see an explosion of sensors and collection of data from the farm” says Eathington.

In spite of the fact that data collection forms a smaller part of ML, it is nonetheless an important precursor evident in any ML application. Agricultural sensors such as ones developed by The Climate Corporation can aggregate lots of data.

Not just that, the company is also providing software models based on ML that has proved better in terms of yield prediction than a manual assessment by soil experts.

One particular research study on ML data says that data collected through sensors will drastically impact ML implementation. It says, “By applying ML to sensor data, farm management systems are evolving into real artificial intelligence systems, providing richer recommendations and insights for the subsequent decisions and actions with the ultimate scope of production improvement.”

Yield Prediction Through ML

The above example illustrates how data and ML can augment agricultural methods. Instances of support vector machines being used to predict soil moisture from remote sensing data to ensemble methods for wholly ascertaining soil information, show how ML can go a long way in assessing different factors related to the agro field.

Coming to predicting the yield of crops, certain studies explore advanced areas of ML such as computer vision. One research demonstrated the automatic counting of coffee fruits through CV at different stages of the coffee plants considered in the study. This way it would give details such as coffee maturation amongst many others. Farmers then get to know the yield and can plan their farming work accordingly.

Another study used SVM algorithm to identify immature citrus fruits. This SVM algorithm consisted of two elements, shape analysis and texture classification. With a testing accuracy of 80 percent, it can certainly help farmers estimate yield for the fruits.

It can easily be seen that ML can vastly optimise different factors in agriculture. With earth’s resources depleting, ML will definitely help cut down on excessive requirements in agriculture. For example, we mentioned disease detection as a factor in estimating yield. Now, if diseases in crop are observed through ML, it can help farmers reduce cost on labor for harvesting. In fact, on a smaller scale, it may help avoid additional labour altogether.

Steady Implementations In India

The true answer to problems in agriculture lies in effective implementations of ML. In countries like India where agriculture is the mainstay occupation, ML can bring in a paradigm shift. With digitisation prevailing in most rural areas in India, ML and AI-related applications are slowly being witnessed in this field.

Earlier in 2016, tech giant Microsoft tied up with International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) and developed a ML-powered mobile application for sowing. This app lets farmers know when to sow through a text message. Dubbed the ‘AI Sowing App’, the technology mainly uses cloud ML and advanced analytics.

Many farmers from regions of Karnataka and Andhra Pradesh, who used this app saw yields growing 30 percent higher. In addition, it also notifies them the risk of pest attacks in the crops which also is an dependent factor for yield.

Interestingly, there are also startups in India that are venturing into agritech space. AgNext and CropIn are two wonderful examples. These companies provide agriculture related solutions by using technologies such as ML and data analytics. Farm management is in fact made simpler with their products.

Although ML-based implementations in agriculture is new in the country, practical uses are like these can encourage more innovations. Conditions like sporadic rainfall, massive floods, pest attacks are strongly evident in the Indian subcontinent. With ML backing agriculture, drastic consequences can be significantly avoided.