HARIT DHARA 5(2) July – December, 2022 Soil Health Management: Knowledge

APPLICATION OF ARTIFICIAL
INTELLIGENCE (AI) IN AGRICULTURE:
AN INDIAN PERSPECTIVE
NISHANT K SINHA, JITENDRA KUMAR, DHIRAJ KUMAR, RAHUL MISHRA, J K THAKUR, K C SHINOGI,
JAYARAMAN SOMASUNDARAM, M MOHANTY

ICAR-Indian Institute of Soil Science, Bhopal, Madhya Pradesh

T he World Government Summit report

published in 2018 Agriculture 4.0 –
The Future of Farming Technology point out
that four key developments viz.,
interaction especially, Artificial Intelligence (AI) is the concept that
can potentially transform the present day agriculture to a ‘produce
more from less inputs’ model. Figure 1 illustrates some uses of AI
in agriculture.
demographics, scarcity of natural resources,
climate change, and food wastage, pressing Analyzing
Market
our future needs from agriculture. Making Demand/
Market Price
use of concepts like big data analytics, Food Managing
Processing Risk
precision agriculture, and Internet of Things
(IoT) to measure agriculture quantitatively
Agriculture 4.0 envision improved crop yield
with less exploitation of environment, low Precision Breeding
Agriculture Seeds
input use, and cost. Agriculture 4.0 AI in
emphasises organization of farm inputs Agriculture
(fertilizers, seeds, farm fuel, and herbicides)
through distributed management practices.
Monitoring
Harvesting
Soil Health
In a scenario where demand for agricultural
produces keeps on increasing along with Feeding
Crops Crop
population growth and changes in lifestyle of (Irrigation/ Protections
Fertilizer)
people, production of enough material using
limited resources is hard without the help of
some innovative concepts. Man-machine Figure 1. Application of AI in agriculture

9
HARIT DHARA 5(2) July – December, 2022 Soil Health Management: Knowledge

At the farm level, AI influences crop production many ways; forecast, type of seeds, and pest infestation in the
particularly through proper distribution of seeds, fertilizers area. Though the recommendation include best
and other agricultural chemicals, automated irrigation choice of crops and technologies for a specific farm,
scheduling, monitoring soil, crop and animal health, it can be further personalized based on the farm's
surveillance of pests and diseases, and farm machinery requirements, local conditions, and past data on
positions in the field. successful farming.

Crop Health Monitoring Systems: Remote

APPLICABILITY OF AI IN AGRICULTURE
sensing techniques, hyperspectral imaging and AI
 Diagnostic Application Service: Identification of build crop health monitoring systems that can
symptoms of water stress, pest and diseases monitor crop health from both time and effort
infestation etc. in farm fields. perspectives. AI facilitates identification of pests,
diseases, and weeds problems and automates the
 Prescriptive Application Service: Soil health
management of these problems. AI-enabled
analysis and prescription of fertilizer recommendation
agricultural production systems also predict future
or any other agricultural inputs
situations and issue advisories for sowing, pest
 Advisory Application Service: Weather advisory control and commodity pricing.
and Irrigation scheduling
Smart Irrigation Systems: Smart irrigation
 Predictive Application Service: Yield prediction,
involves providing the right quantity of water at the
disease and pest attack forecasting (early warning
right place at the right time for the right crop so as
system)
to improve crop yield. Using sensor based
automated irrigation system issues associated with
POTENTIAL OF AI IN INDIAN AGRICULTURE the low irrigation efficiency of Indian agriculture
Agricultural Growth driven by IoT: AI and IoT (sensors) (around 38%) can be resolved to a greater extent.
Also, automated irrigation scheduling is possible
can be used to create intelligent systems that can be
embedded in machines to make it work with higher using machines trained on historical weather
accuracy. These technologies help in the spatial and patterns and soil quality of the locality along with
temporal evaluation of individual plots or plants. the kind of crops to be grown.

Image-based Insight Generation: Using the satellite Agriculture Risk Management: Climate change
and drone imagery real-time alerts can be generated in issues such as increased temperature, erratic
precision farming. These artificial intelligence systems not rainfall pattern and other associated problems have
only save time but also increase safety and reduce increased the importance of the AI system in
potential human error while improving effectiveness. agriculture in recent years. AI based technologies
support farmers in the management of risk and
Computer Vision: Computer vision technology can be uncertainty in agriculture by facilitating the
used to grade the agricultural produce (grading of fruits and preparedness of farmers to handle the crisis
vegetables), enhance the quality of produce, and fetch high efficiently.
market values and profit to farmers. This will help to reduce
post-harvest losses of perishable commodities. Also, the
Natural Language Processing (NLP) for Agro-
technology can make use in the procurement, storage, and Advisory: India is a multi-lingual society and
distribution of food grains. majority of farmers are illiterate. A lot of content
failed to reach the desired people due to lack of
Identification of Optimal Agronomic Product Mix: AI human-resource to convert it to the end-user's
helps to generate farm specific recommendations language. This gap can be effectively filled through
analyzing different parameters like soil health, weather natural language processing.

10
HARIT DHARA 5(2) July – December, 2022 Soil Health Management: Knowledge

Attract Youth to Agriculture: Decreasing agriculture FUTURE OF AI IN INDIA

labour force due to migration of youth to other jobs can be
Applications of AI based tools in agriculture have
well addressed using AI based agriculture. AI based
initiated in India by several start-ups working in this
technologies will attract the tech-savvy youth, reduce
area to help farmers with improved productivity and
drudgery and save natural resources and agriculture
profitability from agriculture. India’s burgeoning
start-up ecosystem has been actively playing its
HURDLES IN THE ADOPTION OF AI IN INDIAN part in developing the agriculture sector. Since,
AGRICULTURE opportunity in agritech exists across the value chain
from improving farmers’ access to markets, inputs,
Policy: Incomprehensive data governance and data rights
data, advisory, credit and insurance; India can
regime, and lack of enforcement of data regulations,
tackle the issues associated with adoption of AI
privacy, and transparency.
based technologies by providing a suitable
Culture and Society: Barriers were recognized and ecosystem to these start-ups to access the data
prioritized i.e., risk-aversion and resistance to change, lack and market. Moreover, National Strategy for
of trust in technology, and insufficient support of Artificial Intelligence released by NITI Aayog in
universities in data digitization and digital agriculture. June 2018 identifies agriculture as one of the focus
areas.
Education and Skill: factors identified as inhibitors of AI
adoption include language barrier, high illiteracy rates, and To maximize farm output from limited resources
the digital divide, lack of formal, non-formal, and informal agriculture in India should to make use of
education in data engineering, data analysis, data science sophisticated deep techs in future. Deep-Tech
and insufficient proficiency. innovations support farmers to grow crops even in
arid areas with high resource use efficiency, using
Information and Communications Technology (ICT) technologies like AI and ML, robots, temperature
and Data Infrastructure: Inhibiting factors related to ICT and moisture sensors, aerial images, and GPS. An
and data infrastructure include lack of supporting ICT and important aspect of AI is the system's response
data infrastructure (data collection, transmission, storage, time and accuracy. Even behavioral changes in
processing, cleaning, and analysis devices, missing field crops due to changes in microclimate
historical data and insufficient digitization and labeling of conditions can be analyzed in the quickest
data), deficient telecommunication networks and poor response time with accurate information. However,
internet connectivity, low bandwidth and slow network concerns about durability of AI technologies may
performance, limited access to cloud-hosted data, irregular discourage farmers from its adoption as
and erratic electricity supply, fragmentation of data and technologies are changing very fast in this digital
lack of data standards. era, and changing devices and sensors quickly with
advancement in the adopted technologies is not
Finance and Investing: Barriers related to finance and going to be economical for the small-scale farmers
investing are insufficient capital to invest in ICT devices of India. It is certain that digital innovation can
and data infrastructures, fund deficiency for the transform Indian agriculture if there are proper
maintenance of existing infrastructures, lack of public efforts to convince the vend users about the
investments to bridge gaps in data engineering, data potential of AI based technologies in agriculture
analysis, and data science education, low awareness and sector; not only from the user and consumer side
clarity regarding return on investment in AI systems, and but also to the governance and policy side.
no financial assistance schemes for small farms to adopt
and deploy ICT devices and embedded systems. ***

11