The world is progressing with all sort of catastrophic events facing the wrath of nature. The sudden outbreak of such dangerous pandemic COVID 19 has shaken the whole world.No sector has escaped its impact. Its impact on the agricultural sector is complex and varied across diverse segments that form the agricultural value chain. Despite many exemptions from lockdown, the agricultural sector in India has experienced the impetus of COVID crisis. Agriculture in India employs about 55% of the population and contributes 17% of the gross domestic product. Ergo, the functioning of the agricultural supply chain are necessary for the food and nutritional security of India.
The FAO projections indicate that world food demand may increase by 70% by 2050. Moreover, the world population will grow by 35%, from 6.9 billion in 2010 to 9.3 billion in 2050. To feed the growing population the question arises how can India and the governments, in particular, intervene in agricultural value chains to help cope with the shocks caused by the coronavirus pandemic?
Agriculture is at the forefront of the leveraging new technologies and these technologies are in the thirst of gathering data. Now agriculture is in the phase of open,data-driven collaboration and digitally-enabled collective action for recovering and building a resilient global food system. Therefore, considering the humungous implications on and barriers in the field of agriculture, Big data comes into action. Some of the problems such as –
- Limited arable land,
- Receding of the water level,
- Training on the use of the devices,
- Basic troubleshooting,
- Use of data, use of smartphones and app, and more. The concerns don’t end there. Problems like infrastructure, need for uninterrupted power and internet connectivity, and finance to deploy the technology are always a concern.
Traditionally, agriculture was solely dependent upon historical data of soil, land, geography and weather forecasts which were in completely unstructured and scattered format. In some areas digital innovation was minimal But to meet such challenges there has been a growing interest in the field of large-scale data analytics in recent years. Many Agritech companies are in the race of building and enabling the platform to serve the purpose. With advances in processing speeds, the availability of cheap cloud storage, and log data, data can be extracted and analysed to obtain accurate results quickly.
Today, the way we look at Artificial Intelligence (AI) and Machine Learning (ML) is completely redefined with the new age methodologies and devices interacting with Cloud Services.
Some of the areas where we see innovations using AI, ML and IoT especially in farming are:
- Best crops to grow, predictions on weather/rains
- Recognize crop diseases
- Use of minimal resources such as water and managing soil quality
- Responding to emergencies
Here is the high-level diagram showing how different parts and scenarios of agriculture ecosystem can leverage cloud and AI services.
MICROSOFT HAS COME UP WITH AN AZURE CLOUD SERVICE CALLED AURE FARM BEATS DURING THE MICROSOFT IGNITE EVENT
Microsoft is democratizing agricultural intelligence providing data-driven solutions in a simple and affordable manner. Azure FarmBeats, startups can aggregate agricultural data from different sources, build AI/ML models using the data fused from sensors, drones and satellites, and also set-up their own customized digital agriculture solution. Thus, the Azure market place is a multi-year effort to bring robust data analytics to the agriculture sector. The International Food Policy Research Institute claims these can boost farm productivity by as much as 67% while reducing resource usage.
According to the Traxn data, there are 535 agritech startups in India while as per Nasscom every ninth agritech startup in the world is from India. These startups are tapping into new areas such as Market linkages, digital agriculture, improved access to farm inputs,farming-as-a-service etc. Cropln, Agrostar, Crofarms, Bijak, Intello labs, Gramaphone are among the leading startups in the agritech segment in India. The sector is growing at 25 per cent year-on-year rate.
The programme — Microsoft for Agritech Startups — will provide young enterprises access to Microsoft’s cloud platform to help them innovate and scale fast. According to Microsoft, startups can also get access to Azure FarmBeats, which can help them focus on core value-adds instead of the undifferentiated heavy lifting of data engineering.
The FAO estimates find that: there are at least 570 million farms worldwide, of which more than 500 million can be considered family farms. Most of the world’s farms are small, with more than 475 million farms being less than 2 hectares in size. The techniques like precision agriculture and phenotype mapping are a very efficient method of improving yield, ensuring sustainability, reduction of cost in a farmer’s field. Therefore, given that the befits are known do you think a farmer can afford such an expensive technique? According to USDA, the high cost of data collection prevents farmers from using data-driven agriculture. Let me sight one example of how costly the affair is-In an expo of the latest precision Agri equipment, the cheapest sensors that were available there were five sensors for $ 8000 and a recurring cost. Thus, seeing such a huge involvement of cost, small and marginal farmers would be reluctant in the adoption of these techniques. As a result, the Microsoft farm beats came into the gameplay, where the goal is to bring down the cost of these data-driven agricultural solutions by two orders of magnitude i.e. from $8000 to $80.
In the rural area there a lot of unused spectrum (hundreds of megabits per seconds of unused capacity at the point which sensors, cameras, drones, tractors can be fused to extract a lot of information from the middle of the farm) by utilization of these TV white spaces just like the WIFI connects the houses.
When we need to map the soil moisture level 6 inches below the soil, we need a lot of sensors to get deployed on the farm field for building an accurate map. And we need to put sensors in every 10m. But the problem arises as the sensors are awfully expensive and it could also be an interruption in the farming activity. Therefore, can we build a map using a few sensors? It can be done by Use of UAVs to enhance spatial coverage which can help in covering visual data with sensor data from the farm.
TEATHERED HELIUM BALOONS(TYE)
These are the aerial imagery collected for building the maps.
a) Orthomosaic-A 40 MPixel orthomosaic created from a 3-minute flight over 2-acre area of the farm.
b) The predicted soil moisture map (our sensors measure moisture on a scale of 1 to 5). Note that the top-left region in the image where the ground appears wet was correctly predicted to have high moisture even though no moisture sensors were present in that part of the farm.
(c) The predicted pH map (pH is measured from 0-14, 7 is neutral and 0 is the most acidic). The system identified that the whole field is slightly acidic, but the bottom left/centre is more acidic than the rest.
(d) The predicted soil temperature map (in Fahrenheit scale).
A vibrant agricultural sector has been the basis for a successful economy. Therefore, proper connected digital farming will be able to make better informed and long-lasting decisions. Finally, new and disruptive business models will enable the data-driven agri-food chain. As a result, we need to work on data-centric issues, while also finding new solutions to classical problems of using IOT devices and wireless communication in a harsh and rural environment. However, the sustainability of IoT based business depends on both supply (providers of IoT technology) and demand(agri-food users) stakeholders. This is in context of large-scale deployments of resources to the end-users and validating the related benefits.