top of page

AI implementation in Indian Agriculture

Edited By: Lavanya Goswami


The diversity of cultivation styles has led the Indian agricultural environment to be different in comparison to other countries. 75% of the current agricultural land is devoted to farming food crops like wheat, rice and millets meanwhile the remaining 25% is for commercial crops. There is a distinct lack of agricultural diversity among marginal farmers (<0.01 hectares of land) in India, with farmers even in rain-fed areas sticking to food crops when they could shift to other commodities and earn a higher profit. Crop selection by a majority of the farmers is done on the basis of procurement of low-cost seeds and essentials, and not the climate, soil moisture and quality. In parts of India, the over-pumping of water for agricultural use is leading to falling groundwater levels. Conversely, water-logging is leading to the build-up of salts in the soils of some irrigated areas. In rain-fed areas on the other hand, where the majority of the rural population live, agricultural practices need adapting to reduce soil erosion and increase the absorption of rainfall.

India is a heavily populated country, which means that agriculture is still quite labor intensive. Overpopulation in the sector has also led to the land being fragmented with the average land size being just above 1.08 ha.





SOURCE: AGRICULTURAL CENSUS 2015-16

This has adversely reduced productivity, with India producing 3479 kgs wheat per hectare. Comparatively, France produces 7171 kgs per hectare and Britain 6967 kgs per hectare. India has fallen behind countries with less arable land in terms of production– in fact, India has only one-third of China’s land productivity, which is a cause for concern.





SOURCE: AGRICULTURAL CENSUS 2015-16

The majority of Indian farmers (68.45%) have marginal size holdings, which means that their capacity to earn and invest into the land is restricted due to income and productivity constraints.

These problems combine to induce poverty among agricultural households. Small land sizes with less productivity does not give them much options for profit making, with most of the farmers earning the bare minimum. This makes investment in upgraded technology difficult as farmers don’t make enough money to invest in production augmenting technology which is one of the reasons why a majority of farmers in India still stick to traditional methods of farming, which results in lesser land productivity and a higher burden on the country’s natural resources.



Introduction to AI

Artificial intelligence, shortened as AI, refers to the ability of machines to perform tasks that normally require human intelligence. AI is defined by a set of capabilities, rather than a specific technical approach to achieving those capabilities.

Machine Learning can be understood as a subset of AI, a technology that uses unique algorithms to resolve and sort data, and in the process, learns more about the data and comes up with a prediction or suggestion about the data. When we say “incorporate AI in agriculture”, it basically means a combination of Machine Learning and IoT– Machine learning (ML) allows the Internet of Things (IoT) to gain hidden insights from the treasure trove of sensed data and be truly ubiquitous without explicitly looking for knowledge and data patterns.


The use of IoT and AI technologies has the potential to result in a positive transformation of traditional agriculture, by improving the use of data collected from smart agriculture sensors, managing and governing the internal processes within the smart agriculture environment (including the management of the harvesting and storage of crops), waste reduction and cost saving, increasing business efficiency by means of automating traditional processes, and improving the quality and volume of products. Traditional Indian agriculture lacks the system of providing correct information to farmers at the right time, which can be resolved using IoT and AI technology.


Agriculture and AI

There are many IoT applications in the agricultural sector – the advanced agriculture practices based on technology of information is called smart agriculture. Smart farming or smart agriculture is a kind of agriculture, in which several modern technologies are used in both agricultural and livestock sectors in order to increase the quantity of production, and its quality through maximizing usage of the resources and minimizing the environmental impacts. This is done via soil management, automated irrigation and fertilization, pre-stage disease detection, crop scheduling and weed management.

Adoption of smart farming techniques powered by IoT has several positive implications for the development of the Indian agricultural sector. These techniques have the potential to increase land productivity, ensure efficient distribution of resources, minimize crop wastage and improve labor productivity as well. However, the traditional method of farming still followed in India makes it difficult to smoothly incorporate such a system, unlike other developed countries like the USA and South Korea. If it is to be implemented, it has to start from an experimentation phase, which requires investment and government supervision.


Indian agro digitalization on the production end – challenges and the way forward

Application of Artificial Intelligence in today’s nature of Indian agriculture would be an ambitious and overreaching project. Indian agriculture still follows traditional means of production and in most areas, farmers fail to utilize existing technologies that have actually become obsolete now in other developed countries. However, if implemented, AI could bring about drastic changes in the productivity and food security of the sector. It could be the key to solving the main issues challenging Indian agriculture such as improper irrigation, decreasing soil fertility, lack of crop diversity and crop waste due to late disease detection. There are challenges to AI incorporation in India that are quite systemic to the agricultural system itself that need to be solved if we are to go forward.


Lack of big data is one of the most prominent features of the challenges facing AI implementation. IoT cannot work without big data, and most Indian farmers are rural based and do not have the resources or knowledge to register the details of their farming. The Indian system is also not comprehensive enough to maintain systemic records of each individual farmer, their output, number of pesticides used, intervals of timely irrigation, diseases detected, etc. AI relies on this data at a large scale in order to make accurate predictions and the lack of such data would render AI applications useless. The collection of big data is also a time-consuming process as harvest seasons are irregular in India and it would be impossible to implement AI tech immediately, especially programs like disease detection, soil management and yield prediction. However, the process of digitization of land records is underway in India, which could be a starting step to build big data for IoT.


Small land size and fragmented holding are also another barrier for smooth implementation. Due to large family sizes in rural India and its inheritance process, land holding gets more and more fragmented by the generation which further puts a burden on its productivity. As most rural households practicing agriculture are low-income in nature, they would not be able to invest a large amount of money for AI applications on small land holdings which won’t give them the required yield to pay back loans.


As shown below, credit availing facilities are skewed in the favor of large-scale farmers, who have better access to institutional credit, leaving the marginal farmers who form the majority, at a disadvantage.





SOURCE: NSS 77TH ROUND

The average monthly rural household income in India stands at 10,218 rupees (NSS, 2019). However, excluding large scale farmers, the average household income stands at a low 8,236.75 rupees. This does not allow for the time or risk factor allowed in investing in AI technology. The farmers are risk averse and wary of new technology which makes it a socio-economic barrier to technology integration.


Lack of technology knowledge and access also restricts the scope of application. Data from the report 'Internet in India Report 2022' shows that of the total 759 million, nearly 399 million users are from rural India, while urban India has around 360 million users. According to the study, the addition of users has slowed down since 2020 to sub-15 percent (Forbes, 2023). However, this does not guarantee the knowledge of technology to implement AI in agricultural systems. IoT usage for smart irrigation depends on interconnection with wi-fi, which is still not universal in India that depends on smartphone data. IoT applications require the placement of sensors and sometimes even make use of drone technology to gather data on the soil and crop systems, which might not be easily accepted by farmers unwilling to farm over sensors.


The persistent fear of AI replacing human jobs has also taken over the mindsets of people in India. When one hears the term AI, mechanization and robotics are the main components that come to mind. The development of AI for agriculture is mainly being done to reduce agricultural dependence on labor intensive production and switch to data driven analytic production. However, the lack of an educating body in India and inaccessibility of rural farmers to technology driven development would make AI application in India difficult.

The irrigation system in India is also not as well-connected as in developed countries. Most countries have an interconnected irrigation system that is linked to the nearest reservoir or water bank. However, the Indian system is disjointed– some farmers depend solely on rainwater whereas others use wells for groundwater irrigation. This makes it difficult to digitize the irrigation canals, especially when some of them also lack proper cementation. Infrastructural overhaul would be needed if AI is to be implemented in a comprehensive manner, even on the small-scale level. It would also require a third-party overseeing body as farmers would be unaccustomed to the technology.


Implementing AI in farming would also have adverse effects on the culture surrounding farming in India. Smart farming systems in South Korea were specifically designed to integrate the rural economy and make sure that they don’t fall behind in Korea’s technology and development expansion. However, if left to private actors, the technological development would be focused at large scale farmers, leaving the small family scale farm businesses behind in terms of technological development. There could be concentration of technology in the hands of large commodity crop producers which causes further divide in the agro sector and has scope of reducing crop diversity by forming monocultures of the most profitable cash crops.


Smart farming technology as a destructive innovation– what happens if India can’t keep up?

Smart Farming techniques that make use of artificial intelligence, big data and the Internet of Things are being utilized at a preliminary stage in countries like the USA, South Korea, China and Japan. It is also being used in automation in European countries. These countries have on the whole less arable land than India, but their per hectare crop productivity is much higher. With that in mind, they already have an advantage over developing countries and the application of AI in their agricultural sector is taking them closer to food security and sustainability.

This is intrinsically problematic as smart farming technology is a destructive innovation– it is taking over labor driven traditional farming methods and replacing them with a data-driven and more scientific approach that maximizes production and is also mindful of sustainability and climate protection. If India fails to catch up, it will fall behind countries in terms of agricultural production, even those with less arable land like South Korea. In terms of paddy and rice production, China, USA and Brazil have overtaken India, even when it is the country with the most arable land in the world. Consistently, these countries are also one of the first implementers of smart farming technology in recent years.

Currently the Indian farming system is stuck in a vicious cycle– the credit giving and information system for farmers is inefficient which poses problems for them in terms of getting funds. Even if they do get loans, the farm sizes are too small to implement personalized smart farming techniques to improve productivity. The availing of loans via government institutions is easier for large scale farmers, which perpetuates income divide.

However, it is important that Indian farming not face the brunt of the destructive innovation that is AI. Since the agricultural sector employs more than half of India’s population, the country cannot afford to fall behind in terms of production and sustainability and in this era of SDGs, food sustainability and environment protection is key.


Conclusion

Artificial Intelligence application in agriculture could be one of the factors that leads to developing nations catching up with the productive capacities of advanced countries. Smart farming technologies are destructive innovations that need to be capitalized on, especially countries like India that face challenges in the production end of the agricultural sector. Letting private sector competition drive smart farming adoption in India would not be the best bet as it would perpetuate the income and technology divide between large scale and small-scale farmers, which is already quite wide. A nation led approach with comprehensive government plans to implement new technologies in agricultural communities to solve issues of soil, crop and disease management would be a good first step. The government should also consider resource and technology sharing plans with other developed countries and work on domestic R&D optimization to integrate start-ups to develop smart farming technologies that are compatible with the Indian agriculture system and climate. AI incorporation in agriculture would be a huge leap forward in terms of sustainable production and climate goals, and if done equitably, would be the much needed financial and technological boost for the small-scale farmers of the country.


Σχόλια


bottom of page