Have you ever wondered, why Agriculture is always a job of risks? Because it is always a job of decision making. Urea or Ammonium nitrate? Urad dal or Vegetable crop? Hand weeding or spraying? Local Mandi or Export? Stock it or sell it?
Remember from farm to fork it has to travel through multiple intermediaries, involving multiple variables and hence multiple decision making at every single level. At the same time, at every single level, enormous data is generated. Hardly these data are collected, stored or processed for decision making. Here comes the role of ‘Big data’ in Agriculture. Big data not only refers to data itself but also set of technologies that capture, store, manage, analyse – large and variable collections of data to solve complex problems. Hence, Big Data is envisaged to solve enduring problems and future menaces in Indian agriculture.
Data generation is the first step to utilize Big data in Agriculture. According to UNECE (2013), there are in general three categories of data generation, 1. Process mediated, 2. Machine generated and 3. Human sourced. Process mediated involves traditional data that is recorded in any farm like purchasing inputs, fertilizer application schedule, yield, etc., It has to be collected from individual farms, then structured and stored in database systems. This could be done in India through wide network of extension workers or agricultural officers at grassroot level.
Second, Machine generated (MG) data, as the name denotes, it is obtained from simple sensor records to complex computer logs. Now a days, MG data is increasing available through adaptation of smart farming and Integration of IoT. Farms using drones, infrared cameras, GPS augmented machines, dairy farms using robots for milking cows are not uncommon in India. These data form an important MG structured information for big data aided farming.
And finally, Human sourced data, these are loosely structured and ungoverned data recorded in books, photos, audio or video format. With rich traditional farm knowledge these datas in India are highly valuable. But integration from diverse source and poor awareness to store the data are the challenges of this Human sourced data in India. Besides the above said options for data generation, Government sources like NSSO survey, farm household survey, Census, Socio Economic Caste census, Meteorological data, land records could also be used as sources of data for ‘Big Data’ administered farming.
“With rich traditional farm knowledge these datas in India are highly valuable. But integration from diverse source and poor awareness to store the data are the challenges of this data in India.”
Applications of ‘Big Data’ in Agriculture
The voluminous data available could be used primarily for crop forecasts. Like, Meteorological data for weather predictions assisting sowing or other intercultural operations. On the other end it is also used for Market forecasts ensuring better prices for farm produce. In a similar way, ‘Farmer Business Network’, an US based company pools data from numerous small farmers and in turn shares the insights to its members. They provide information on yields, supply prices and other informations that help small farmers to compete with larger ones. Likewise Big Data has huge application for crop forecasts for millions of small farmers in India.
Big Data and IoT (Internet of Things) work in conjunction, as the latter serves as a means of gathering data. Now a days IoT has revolutionized the way through which agricultural machineries and farm tools are used for sending and receiving the datas. Precision farming is exploiting this advantage and through this focus is shifted to per plant productivity rather than productivity per unit of area. Putting this into operation, a start-up ‘AGER point’ in Florida that produces nuts and citrus orchard management software using satellite data. This provides tree specific information and boosts production per plant by accessing their needs individually. As the Indian government dreams for doubling farm income by 2022 in which raising productivity has been given primary impetus, the trio – Big Data, smart farm machineries and IoT would help to achieve the dream.
“As the Indian government dreams for doubling farm income by 2022 in which raising productivity has been given primary impetus, the trio – Big Data, smart farm machineries and IoT would help to achieve the dream.”
Although Big Data provides keen insights into the existing farm business and operations, cloud technology makes it possible to store this data and make it accessible to all. Already cloud technology, a virtual storage space is exploited in agricultural activities with Big Data analytics. Crop X, an Israeli start-up developed a small ball like field sensors integrated with cloud based software. This system automatically delivers required amount of water to each plant instead of whole field at a time. This saves energy and water, the scarce resource of Israel and also to India.
Similarly, Alibaba, China’s largest e-commerce operator is not an exemption to experiment cloud technology enabled agriculture. Alibaba recently launched ‘ET Agricultural Brain’ that uses Alibaba’s cloud services to boost agricultural efficiency, crop yields and income of China’s small farmers. Earlier, Alibaba had used it’s cloud services to raise pig by providing digital ID to each pig. This tracks their move, enhances focus on newborns and has resulted in decrease of death rates thereby raising income by 10 percent. Likewise, Indian Government should also create an enabling environment to private players in research and experiment of Big Data analytics and cloud computing in Indian farms.
Challenges associated with ‘Big Data’ Agriculture
As India steps into Big Data aided farming, it should also consider the challenges and concerns of handling voluminous data. Data generated might get concentrated to very few companies that could be deployed for potent abuses by GMO lobbies, distorting agriculture commodities market and manipulation of agri companies. The other main concern in Indian situation is affordability to farmers. On speaking about smart machines, we should not forget the fact that India with mechanization level 40 – 45 percent still lags behind BRICS countries.
Quality of data is another main concern. We have to go a long way in documentation, as maintenance of registers and data books in traditional farming is nearly negligible in India. My experience as an agriculture graduate in collection of data in surveys often encountered with rough datas and approximate values, making it as ‘dirty’ data with high inaccuracies.
“My experience as an agriculture graduate in collection of data in surveys often encountered with rough datas and approximate values, making it as ‘dirty’ data with high inaccuracies.”
For that, Government should roll the dice to start the game by framing Big Data policy for agriculture. At first Government should open up the data it has and setup a dedicated institute with skilled professionals to facilitate the entire program of ‘Big Data enabled farming’.
Originally published in ‘The Agraria’ e-Magazine.