Future of Precision Agriculture in India

Full text

Future of Precision Agriculture in India
Dipankar Bhattacharyay, Sagar Maitra*, Sandipan Pine, Tanmoy Shankar and
Pedda Ghouse Peera S.K.
Center for Smart Agriculture, Centurion University of Technology and Management, Odisha-761211
*Corresponding author: sagar.maitra@cutm.ac.in
Abstract
The yield and quality of crops depends on various biotic, abiotic and management related
parameters. In conventional agriculture the farmers relied on their experiences. Due to
human perception often there is uncontrolled use of resources and inputs resulting in
not only natural resources wastage and environmental pollution but also nancial loss
of farmers. Precision agriculture uses technology such GPS, sensors, Internet of Things,
robotics, drones, machine learning al decision support systems etc. to optimize the use
of natural resources and farm inputs for a specied yield and quality of crops. The future
of precision farming is moving towards extensive use of machine learning techniques
and image analysis. However, the major constraints are loss of job, data security, lack of
motivation, training and so on.
Keywords: Aotic, abiotic, management, environmental pollution, crops, farming
1. Introducon
Agricultural production is the result of the combined effects of natural resources,
biotic factors, agro inputs and management. A harmony in interaction of above
can only make it a sustainable venture. In the developing countries a considerable
portion of population remains engaged in agriculture as in India about 54 percent
of people are directly and indirectly involved in agriculture and allied activities.
However, agriculture contributes only 15 percent to country’s gross domestic product
(GDP). Indian agriculture passed a long journey since independence. The country
witnessed nightmare of absolute shortage of food grains supplies in the 1960s,
however, achieved sufciency after adoption of Green Revolution. At present, India
is one of leaders of crop producer with a production of 285 million tonnes of food
grains in 2018-19 and estimated to reach at 292 million tonnes in 2019-20 (Economic
32
Chapter
Future of Precision Agriculture in India. In: Protected Cultivation and Smart Agriculture edited by
Sagar Maitra, Dinkar J Gaikwad and Tanmoy Shankar © New Delhi Publishers, New Delhi: 2020, (pp.
289-299). ISBN: 978-81-948993-2-7, D OI: 10.30954/NDP-PCSA.2020.32

290 Protected Cultivation and Smart Agriculture
Times, 2020). India’s horticulture output is around 311 million tonnes in 2018-19
(GOI, 2019). Besides, India is the second largest producer of fruits and vegetables
worldwide. In milk production India ranks rst in the world accounting more than
13 per cent. The agricultural growth rate is supported by favourable investments,
technological development and policy support.
The green revolution is also associated with negative agro-ecological consequences.
During recent period, India is facing a great challenge in terms of achieving much
needed agricultural sustainability to feed the populous nation because of threats
appeared as yield plateauing, degradation of land and water, scarcity of irrigation
water, genetic erosion and vulnerability in farming, climate change and associated
problem and inferior yield harvested by the farmers than the potential and poor
realization of consumer currency. The solution of these problems is complex.
An alternative left for enhancement of agricultural productivity and assuring
sustainability from limited natural resource without any negative consequences is by
efcient utilization of resources and maximization of input use efciency. Now it is
the right time to exploit the modern tools by accumulating all available technologies
into the platform of agriculture for enhancement of return from agriculture as well
as to achieve sustainability in crop production. Precision agriculture (PA) can merge
all technologies with agriculture for boosting up the productivity with input use
efciency (Hakim et al. 2016; Sha et al. 2019).
2. Precision Agriculture and its Scope in India
Green revolution provided self-sufciency in food grains production. In spite of the
signicant development in agriculture, the average productivity of many major crops
are far less than of potential productivity of Indian high yielding varieties/ hybrids
with fullest utilization of resources and inputs. Further, for some of the crops, the
average yield output in India is less than other countries with existing technologies.
Interestingly, the crop yield of the most agriculturally rich locality of the country is
far below than the average yield of many high productive countries of the world (Ray
et al. 2001). Moreover, the environmental issues associated with agriculture may be
considered as a warning against over exploitation of resources. These factors warrant
the need for application of latest, improved and inter-disciplinary approach as PA.
PA is a data and technology-driven farming practice to detect, analyze and take
appropriate measures to manage the variations in parameters of the eld. Its goal is
to optimize productivity and protability, and ensure protection of natural resources
and sustainability. In PA, new algorithms are used to improve the decision making
process for managing different aspects of crop production. Advancement in space
technology and IT revolution created a new replacement arena for farm sectors.
Therefore under the changed circumstances it is essential to understand over the
new cutting edge technologies evolved in the eld of agriculture and customized
modication as per the farm sectors. It is true that at present pan India adoption of PA

Future of Precision Agriculture in India 291
is tough as because entire farm sector is not ready to adopt sophisticated and rened
technology, however, there are some comparatively developed pockets as well as
activities managed by progressive farmers where PA can be adopted as incubators for
improved technologies.
Agriculture in India is performed mainly by small and marginal farmers under the
close supervision of farmer’s family members and farming is a family responsibility
to them. These small and marginal farmers of India mainly adopt management,
apply input and take decision related to farming on the basis of their knowledge
and experience, nancial capacity, extension services provided and availability of
resources. Micro-situation specic variability agricultural practices are commonly
prevailing in the developing countries in decision making by the farmers and this
close relation of the farmers with crop management may be considered as some sort
of spatial treatment based on supervision and experience. But this crude form of
precision technology is needed ne tuning and modication in the light of modern
technologies (Mondal and Basu, 2009). Recent researches in integration farmer
knowledge, precision agriculture tools and crop simulation modeling can be very
useful for country like India in taking decision and management for poor-performing
patches in farming (Oliver et al. 2010). In this way India can enter into the eld of
evergreen revolution.
3. Need for Precision Agriculture
Agriculture of the developing countries as well as the global food system is facing
many challenges today and technocrats are suggesting some management options
with existing and proven technologies. But under ever-changing environmental
conditions it will be difcult to cope up with future challenges which may require
involvement of novel technology based approaches. The reduction in productivity
per unit area, decreasing and degrading natural resources, emerging threat due to
global warming and climate change and stagnation in farm income have posed a
major challenge in agricultural prosperity. These ultimately direct for adoption of the
newly developed technology options for sustaining farm productivity. Generally, an
entire eld is managed by adoption of recommended package of practices developed
on the basis of some average condition, which may or may not exist in the entire
farmland, therefore, precise crop management is needed which can recognize
site-specic variables within agricultural lands and adjust management strategies
accordingly with a better capability of decision making. The progressive farmers
are aware that the variability in yields across the landscape and many times they
manipulate management practices as per their previous experience. These variations
can be traced scientically by different tools for better management practices in terms
of responsiveness to different yield causing factors. Precision agriculture provides
the scope to automate the collection and analysis of data for accuracy to make an
appropriate decision.

292 Protected Cultivation and Smart Agriculture
4. Convenonal Agriculture vs. Precision Agriculture
The major steps in an agricultural process involve: (a) selecting a location, (b)
soil preparation, (c) seeding and planting, (d) irrigation (e) fertilizer and pesticide
applications, (f) weeding and (g) harvesting. In conventional agriculture utilization
of resources and inputs in all these stages are not optimum, usually on the higher
side, sometimes even on the lower side. This affects adversely the quality of crop,
yield, and environment and above all, the nancial gain of the farmers. This is where
precision farming comes in. It uses technology to determine and deliver the optimum
amount targeting a specied yield.
In place of manually selecting a location people are using GPS or GIS data. Soil
preparation technique based on past experiences of the farmers has been replaced
by use of sensors for measurement of temperature, humidity, volatile matter etc. The
quality of levelling of lands by bullocks and tractors has been improved by utilizing
Laser-guided precision land leveler. Instead of doing seeding and planting manually,
farmers are exploring automated tools such as precision drills, seed drills, broadcast
seeders, air seeders and so on. In place of conventional irrigation system farmers
are using automated and controlled fertigation system empowered by Internet of
Things (IoT) (Maitra et al. 2020). Automated weeding machines and drones for
weed removal and localized application of herbicides have replaced the conventional
manual ways. Mechanical tools and robotic arms have paved their way in place of
manual harvesting. Thus, precision farming shows a marked shift from conventional
one in terms of use of satellite data, drones, sensors, IoT, robotics and precision tools,
the objective being optimization of resources and inputs based on technical data.
5. Components of Variability
The basic steps in precision farming are assessment and management of variability,
followed by evaluation. Information or database is the primary thing in assessing
variability in agriculture. To manage in-elds variability, spatially or temporally, data
related to biotic and abiotic factors are important and databases in this regard need
to be developed. Following are the components where variability can be assessed for
adoption of PA.
Soil: physical properties (texture, structure, moisture holding capacity, bulk
and particle density); chemical properties (pH, electrical conductivity, available
plant nutrients);
Crop: planting geometry (row to row and plant to plant spacing, plant stand);
nutrient composition of standing plant, plant stress due to biotic and abiotic
factors, weed, insect and disease, potential economic and biological yield;
Climate: air temperature, temperature around plant canopy, relative humidity,
rainfall, solar radiation, day length, wind velocity.

Future of Precision Agriculture in India 293
The available multi-disciplinary and latest technologies help in the understanding of
the variations and site specic agronomic recommendations to manage the production
system.
Variability assessment is the most important part in precision agriculture as because
above factors and the processes control the crop performance and yield behaviour and
may change in space and time. Quantication of variability as well as determination
of various combinations responsible for the spatial and temporal differences in crop
productivity is the important challenges for PA. Different scientic technologies
are available for assessing spatial variability and those are applied in PA. Once
the variations are properly understood, requirement of agronomic inputs can be
matched to known conditions on the basis of crop management decisions. Those
are site specic management and use correct applications management equipment.
Enabling technologies can make precision agriculture feasible, economically viable
and sustainable for enhancement of crop productivity. Precision crop management
will certainly enhance input use efciency and greater productivity which will lead to
the ultimate goal for achieving sustainability (Pierce and Nowak, 1999).
6. The Future Shis
6.1 Machine learning applications
Scientists can use articial intelligence/machine learning (ML) based simulations to
evaluate how a type of crop may perform when faced with different soil types, weather
patterns etc. (Mokaya 2019) as well as deliver required amount of nutrient (4R rule)
or pesticide at the right place at the right time. This simulation will help agriculture
scientists to more accurately predict the performance of the crops. Supervised or
unsupervised ML algorithms such as, convolution neural network, Bayesian network,
support vector machine etc. have been used by researchers. Farmers can utilize ML
tools to take appropriate decisions to maximize the return on crops. These applications
are well demonstrated in Fig. 1.
1. Chat-Bot: ML technology can be used to create chat-bots (Mostaço et al. 2018)
to answer to the questions of the farmers, suggesting recommendations on
specic agricultural issues.
2. Unmanned aerial vehicles (UAV): UAV can take pictures and collect data about
a particular location. Its use can reduce operational cost and help monitoring of
a large area. UAV may pave new strategies of improving crop yields through
spraying, counting of plants, detecting anomalies etc. ML techniques can help in
the movement, analysis of data and actions for the UAVs.
3. Robotic agriculture: Robotic agriculture may play a major role in next 10-
15 years. Driverless tractors will be used for farming autonomously. They will
interpret the GPS, radars and sensor data using ML to identify obstacles and
decide the application of the farm inputs (Chunhua and John 2012).

294 Protected Cultivation and Smart Agriculture
4. Automated irrigation system: Automated irrigation system coupled with
conventional weather prediction tools will help to predict the required water
resource. It requires real time ML application to maintain the level of water and
nutrient in soil.
Fig. 1: Machine Learning Applications
5. Crop health monitoring: Monitoring of crop health and exploring possibility of
pest attack can be monitored by combining ML to hyperspectral and multispectral
image analysis. Deep learning ML applications can generate alerts in case of
a disease or pest. Detection at an early stage will help minimize the losses
(Priyanka et al. 2018).
6.2 Big data and IoT
Farmers need to know three things; (1) the parameters which are relatively stable

Future of Precision Agriculture in India 295
during the growing season (2) the parameters showing change and (3) information
to understand why a crop is facing difculty to grow. These three things can be
addressed using big data and Internet of Things (IoT). Analysis of data can help
making appropriate decisions leading to improved yield. The decrease in size of
smart electronic devices will make the decision making process easier.
6.3 Saving of Manual Labour
Precision farming will reduce human intervention. Thus, even on situations,
demanding social distancing precision farming will continue.
6.4 Differential GPS
GPS signals from satellites result in delays to reach the ground while passing through
the different layers of earth’s atmosphere. This time lag causes some errors into the
GPS engine, introducing an error in locating a position. This is commonly known as
pseudo-range errors. Differential GPS basically helps in correcting positional errors
of GPS signals. It uses a xed, known position to eliminate the pseudo-range errors.
A static base station set up on the ground is used to provide correction messages to
the time lags for the signals. In future the trend is moving from GPS to differential
GPS to enhance the precision in position.
6.5 Sensor technology
The sensors used to monitor different parameters are usually costly and are of contact
type. The contact reduces the life of the sensors through chemical interactions.
Thus the future trend is going towards low-cost non-contact type sensors. In this
process there is a major shift towards image analysis-based sensors. However, those
sensors suffer from the effect of lighting conditions. Algorithms have been developed
to minimize the effect of lighting conditions. To meet the need of sensors image
segmentation, feature extraction, deep learning approaches, convolution neural
network (Abdullahi and Zubair 2017) etc. are gaining importance.
7. Constraints
Precision Agriculture is progressing slowly compared to the expected pace. There is
still a need for an advanced decision-support system to make right decisions at the
right time. Less focus on temporal variation, lack of whole-farm focus, precise crop
quality estimation methods, insufcient product tracking and environmental auditing
are hindering the growth of PA in India. Fig. 2 shows possible constraints in precision
agriculture (https://teks.co.in/site/blog/precision-agriculture-top-15-challenges-and-
issues/) may pose challenges in future:
(a) Interoperability of different standards: With the progress in technology
scientists are developing new tools and IoT platforms. The interoperability among
them may create a concern in the future. For success in future, it is necessary to
integrate standalone devices and gateways to holistic, farmer-friendly platforms.

296 Protected Cultivation and Smart Agriculture
(b) Training: PA requires implementation of new-age technologies. For small
farmers, setting up of IoT and sensor networks will become a challenge. Training of
farmers on different PA tools is of signicant importance and the success of PA will
rely on the training. Lack of knowledge can adversely affect the yield or quality.
(c) Internet Connectivity: In many villages strong, reliable internet connectivity is
not available. Unless there is a signicant improvement in network performances and
bandwidth speeds PA will remain problematic. Cloud-based computing also needs to
become stronger. In farmlands with tall, dense trees and/or hilly terrains, reception of
GPS signals can become a major issue.
Fig. 2: Constraints of Precision Agriculture

Future of Precision Agriculture in India 297
(d) Collection of data in agriculture: It modern PA information is gathered from
multiple data points. It is very difcult to monitor and manage every single data point
on a regular basis, over the entire growing seasons. The problem is even signicant
in large multi-crop lands.
(e) Variation in farm production functions: In-depth economic analysis needs
to dene the “correct production function” (output as a function of key inputs, such
as nutrients, fertilizers, irrigation, etc.). However, the production function varies with
crops, various zones of a farm, and also the crop/plant-growth cycle. Unless the
production function is rightly dened, there will always be a chance of application of
inputs in incorrect amounts.
(f) Size of management zones: Traditionally, farmers consider their entire elds as a
single farming unit. That approach will hinder the implementation of PA and increase
the cost of establishing IoT per farmer.
(g) Barriers to entry for small new rms: Due to the signicant cost involvement
in infrastructure as well as upgradation of technology the big players in agro-IoT
will control the PA. It will be difcult for the smaller players to sustain and lack of
competitiveness will be there in future. Lack of competitiveness may increase the
burden on the farmers.
(h) Energy consumption increase: In precision agriculture use of optimum resources
will help move towards a greener planet. On the contrary, use of too many gadgets
may increase the energy requirement. Thus, more resources will be required to meet
the growing need for energy. There is a need to develop gadgets that will consume
less energy for the success of PA in future.
(g) Challenge for indoor farming: Most PA methods suitable for conventional
outdoor farming. Due to the paucity of land farmers are getting involved in vertical
indoor farming. These indoor farming practices are not encouraged by PA.
(h) Technical failures and resultant damages: Signicant dependence of agriculture
on technology may result in serious damage in case of a malfunction. A crop may
face water stress in case of failure of soil moisture sensor. Thus, there is a need to
develop robust sensors and technologies to immediately respond in case of a failure.
(i) Mounting e-wastes: Up-gradation of hardware on a regular basis will lead to
piles of obsolete hardware. Those discarded IoT tools and computers and outdated
electronic devices may create a major problem in future. There is a need to plan the
disposal of the e-wastes.
(j) Loss of manual employment: A large number of agricultural workforces may
lose their job in future. Thus other sectors need to be prepared to absorb the people
who lost their jobs.
(k) Data security: Proper protection of data against malware and data thefts is
required for the success of PA. The PA platform should be ready to prevent attacks
by hackers.

298 Protected Cultivation and Smart Agriculture
(l) Motivation: The positive effect of PA is not felt within a short time. Thus, there
is a need to motivate the farmers otherwise the implementation of PA will become
only a concept.
8. Conclusion
The future of precision farming is mainly extensive application of machine learning
and image analysis techniques. Machine learning techniques will be applied in chat-
bots, unmanned air vehicles, robotics, automated irrigation systems and crop health
monitoring. IoT and big data, differential GPS, non-contact sensors will inuence
precision agriculture in future. Compared to conventional agriculture, precision
farming will optimize the use of resources and inputs based on analysis of acquired
data. Use of optimized resources will be benecial for the farmers as well as for
the environment. However, there will be challenges with respect to loss of job, data
security, sensor malfunction, e-waste handling, interoperability of systems, training,
motivation etc. Proper planning can make precision farming a success in future.
References
[1] Abdullahi, H. and Zubair, O. 2017. Advances of Image Processing in Precision Agriculture:
Using Deep Learning Convolution Neural Network for Soil Nutrient Classication. J.
Multidiscip. Eng. Sci. Technol., 4(8): 7981-7987.
[2] Chunhua, Z. and John, K. 2012. The application of small unmanned aerial systems for
precision agriculture: A review Precision Agriculture, 13: 11119.
[3] Economic Times. https://economictimes.indiatimes.com/news/economy/agriculture/
indias-2019-20-foodgrain-production-to-hit-a-record-high-of-291-95-million-tonnes-
estimates-second-advance-estimate-of-govt/articleshow/74192668.cms?from=mdr;
Accessed on 31 May 2020.
[4] GOI. 2019. Horticultural Statistics at a Glance. Horticulture Statistics Division
Department of Agriculture, Cooperation & Farmers’ Welfare Ministry of Agriculture &
Farmers’ Welfare Government of India, 2018, pp. 458.
[5] Hakkim, V.M.A., Joseph, E.A., Gokul, A.J.A. and Mufeedha, K. 2016. Precision Farming:
The Future of Indian Agriculture. Journal of Applied Biology & Biotechnology, 4(6): 068-
072.
[6] https://teks.co.in/site/blog/precision-agriculture-top-15-challenges-and-issues/
[7] Maitra, S., Shankar, T., Sairam, M. and Pine, S. 2020. Evaluation of Gerbera (Gerbera
jamesonii L.) Cultivars for growth, yield and ower quality under protected cultivation.
Indian Journal of Natural Sciences, 10(60): 20271-20276.
[8] Mokaya, V. 2019. Future of Precision Agriculture in India using Machine learning and
Articial Intelligence, International Journal of Computer Sciences and Engineering,
7(2): 1020-1023.
[9] Mondal, P. and Basu, M. 2009. Adoption of PA Technologies in India and Some
Developing Countries: Scope, Present Status and Strategies. Prog. Nat. Sci., 19: 659-666.
[10] Mostaço, G.M., Ramires, C.S.I.L., Campos, L. and Cugnasca, C.E. 2018. Agronomobot:
A Smart Answering Chatbot Applied To Agricultural Sensor Networks. A paper from the
Proceedings of the 14th International Conference on Precision Agriculture, 14.

Future of Precision Agriculture in India 299
[11] Oliver, Y.M., Robertson, M.J. and Wong, M.T.F. 2010. Integrating farmer knowledge,
precision agriculture tools, and crop simulation modelling to evaluate management
options for poor performing patches in cropping elds. Eur. J. Agron., 32(1): 40-50.
[12] Pierce, J.F. and Nowak, P. 1999. Aspects of precision Agriculture. Advances in Agronomy,
67: 1-85.
[13] Priyanka, R.R., Mahesh, M., Pallavi, S.S., Jayapala, G. and Pooja, M.R. 2018. crop
protection by an alert based system using deep learning concept Research Paper, Isroset-
Journal (IJSRCSE), 6(6): 47-49.
[14] Ray, S.S., Panigrahy, P. and Parihar, J.S. 2001. Role of Remote Sensing for precision
farming – with special Reference to Indian Situation Scientic Note, SAC/RESA/ARG/
AMD/SN/01/2001, Space Applications Center (ISRO), Ahmedabad, pp. 1-21.
[15] Sha, U., Mumtaz, R., García-Nieto, J., Hassan, S.A., Zaidi, S.A.R. and Iqbal, N. 2019.
Precision Agriculture Techniques and Practices: From Considerations to Applications.
Sensors, 19: 3796.