Artificial Intelligence in Agriculture

Global population is expected to reach more than nine billion by 2050 which will require an
increase in agricultural production by 70% in order to fulfil the demand. Only about 10% of
this increased production may come from availability of unused lands and rest of 90% should
be fulfilled by intensification of current production. In this context, use of latest technological
solutions to make farming more efficient, remains one of the greatest necessities. Present
strategies to intensify agricultural production require high energy inputs and market demands
high quality food. The scarcity & increasing labour costs, raising cost of cultivation and crop
failures associated with unpredictable yield due to diseases, failure in rainfall, climatic
variations and loss of soil fertility, fluctuating market price in agriculture commodities etc.,
has made significant negative impact on the socio-economic status on this backbone
population. On the other side the raise in population has created more demand on food grains
resulting with inflation in agriculture commodity prices. Using artificial intelligence we can
develop smart farming practices to minimize loss of farmers and provide them with high
yield. Using artificial intelligence platforms, one can gather large amount of data from
government and public websites or real time monitoring of various data is also possible by
using IoT (Internet of Things) and then can be analysed with accuracy to enable the farmers
for addressing all the uncertain issues faced by farmers in the agriculture sector. By the 2050,
the UN projects that two-thirds of the world’s population will live in urban areas, reducing
the rural workforce. New technologies will be needed to ease the workload on farmers:
Operations will be done remotely, processes will be automated, risks will be identified, and
issues solved. In the future, a farmer’s skills will increasingly be a mix of technology and
biology skills rather than pure agricultural.
Importance of Artificial Intelligence in Agriculture
Artificial Intelligence (AI) can be applied cross disciplinary and it can also bring a paradigm
shift in how we see farming today. AI-powered solutions will not only enable farmers to do
more with less, it will also improve quality and ensure faster go-to-market for crops. Today’s
technology advancement in Artificial Intelligence, Big Data, IoT are becoming the major
drivers for providing the Digital IT solution almost in all the fields and business sectors.
Hence, it is proposed to make use of Digital solution aided with Artificial intelligence to
uplift the habitat of the trampled farmer community while providing yet a new opportunity
for business and entrepreneurs by enabling smart farm as a service.

1. Growth driven by Internet of Things (IOT): Digital transformation is disrupting the

agricultural world. IoT technologies allow correlations of structured and unstructured
data to provide insights into food production. Huge volumes of data get generated every
day in both structured and unstructured format. These relate to data on historical weather
pattern, soil reports, new research, rainfall, pest infestation, images from Drones and
cameras and so on. Cognitive IoT solutions can sense all this data and provide strong
insights to improve yield. Proximity Sensing and Remote Sensing are two technologies
which are primarily used for intelligent data fusion. One use case of this high-resolution
data is Soil Testing. While remote sensing requires sensors to be built into airborne or
satellite systems, proximity sensing requires sensors in contact with soil or at a very
close range. This helps in soil characterization based on the soil below the surface in a
particular place. Hardware solutions like Robot (pertaining to corns) are already pairing
data-collecting software with robotics to prepare the best fertilizer for growing corns in
addition to other activities to maximize output. The IoT enabled sensors need to be
installed in the field at the prescribed locations. These sensors are the transducers that
collect the data on climatic condition, soil moisture & fertility, root & shoot growth;
profused leaves growth, photo-period monitoring, floral & seed setting, grain/fruit
bearing, pest & deceases as critical growth factors symptoms and harvest readiness.
The IoT device includes the transducer that probes the various parameters of
environment and crop mentioned above. It can be mounted on protected mini board with
WiFi device, microcontroller, low cost VGA image sensor, mini battery powered with
micro solar panel. The data can be collected at required time intervals either by installing
WiFi active hot spot towers as required for entire field coverage. Alternatively, drones
 with active WiFi hot spot can also be used to scan and collect data from IoT devices as
well as to capture elevated motion picture of the entire field.

Data-driven farming: By analyzing and correlating information about weather, types of

seeds, soil quality and probability of diseases, historical data, marketplace trends, and
prices, farmers will make more informed decisions.
2. Image-based insight generation: Precision farming is one of the most discussed areas
in farming today. Drone-based images can help in in-depth field analysis, crop
monitoring, scanning of fields and so on. Computer vision technology, IOT and drone
data can be combined to ensure rapid actions by farmers. Feeds from drone image data
can generate alerts in real time to accelerate precision farming. Companies like Aerial
tronics have implemented IBM Watson IoT Platform and the Visual Recognition APIs in
commercial drones for image analysis in real time. Given below are some areas where
computer vision technology can be put to use:

 Disease detection: Pre-processing of image ensure the leaf images are segmented into
areas like background, non-diseased part and diseased part. The diseased part is then
cropped and sends to remote labs for further diagnosis. It also helps in pest identification,
nutrient deficiency recognition and more.

 Crop readiness identification: Images of different crops under white/UV-A light are
captured to determine how ripe the green fruits are. Farmers can create different levels of
readiness based on the crop/fruit category and add them into separate stacks before
sending them to the market.

 Field management: Using high-definition images from airborne systems (drone or

copters), real-time estimates can be made during cultivation period by creating a field map
and identifying areas where crops require water, fertilizer or pesticides. This helps in
resource optimization to a huge extent.

3. Identification of optimal mix for agronomic products

Based on multiple parameters like soil condition, weather forecast, type of seeds and
infestation in a certain area and so on, cognitive solutions make recommendations to farmers
on the best choice of crops and hybrid seeds. The recommendation can be further
personalized based on the farm’s requirement, local conditions and data about successful
farming in the past.
External factors like marketplace trends, prices or consumer needs may also be factored into
enable farmers take a well-informed decision.

4. Health monitoring of crops

Remote sensing techniques along with hyper spectral imaging and 3d laser scanning are
essential to build crop metrics across thousands of acres. It has the potential to bring in a
revolutionary change in terms of how farmlands are monitored by farmers both from time and
effort perspective. This technology will also be used to monitor crops along their entire
lifecycle including report generation in case of anomalies.

5. Automation techniques in irrigation and enabling farmers

In terms of human intensive processes in farming, irrigation is one such process. Machines
trained on historical weather pattern, soil quality and kind of crops to be grown, can automate
irrigation and increase overall yield. With close to 70% of the world’s fresh water being used
in irrigation, automation can help farmers better manage their water problems.

6. Drone Based Technology

One of the most promising areas is agriculture, where drones have the potential to address
major challenges. Drone technology is giving agriculture a high-tech makeover. Here are six
ways drones will be used throughout the crop cycle:

 Soil and field analysis: By producing precise 3-D maps for early soil analysis, drones
can play a role in planning seed planting and gathering data for managing irrigation
and nitrogen levels.

 Planting: Startups have created drone-planting systems that decrease planting costs
by 85 percent. These systems shoot pods with seeds and nutrients into the soil,
providing all the nutrients necessary for growing crops.

 Crop spraying: Drones can scan the ground, spraying in real time for even coverage.
The result: aerial spraying is five times faster with drones than traditional machinery.
 Crop monitoring: Inefficient crop monitoring is a huge obstacle. With drones, time-
series animations can show the development of a crop and reveal production
inefficiencies, enabling better management.
 Irrigation: Sensor drones can identify which parts of a field are dry or need
improvement.

 Health assessment: By scanning a crop using both visible and near-infrared light,
drone-carried devices can help track changes in plants and indicate their health and
alert farmers to disease. UAVs may one day consist of autonomous swarms of drones,
collecting data and performing tasks. The biggest obstacle to that becoming a reality
is sensors capable of collecting high-quality data and number crunching software that
can make that high-tech dream a reality.
Models for Farmers Services
The beneficiary of this service can be offered with following service models. (a) Chatbot (b)
Agri-Ecalculator for suitable crop selection along with resource estimation (c) Crop care
service. (d) Price prediction and market guidance (e) Crop loan and insurance service.

 Chatbot: Currently, AI-powered chatbots (virtual assistants) are used in retail, travel,
media, and insurance sectors. But agriculture could also leverage this technology by
assisting farmers with answers and recommendations on specific problems. This
service lets the farmer to get their queries answered via interactive voice chat in their
native languages. The chatbot engine is driven with both supervised and reinforced
machine learning techniques for a continuous and context sensitive learning. Thereby
the chatbot answers to most of the generic queries before it lets to human operator
intervention for any queries that are unique in nature.

 Agri-E-Calculator: The agri-e-calculator as a smart application help the smart

farmer to choose the most suitable crop and affordability based on several dependency
factors. The farmer can use the smart calculator and just choose the desired crop to be
cultivated over his preferred coverage area of farm. Then all other required inputs
based on various dependency factors are automatically identified and taken by the
e-calculator and provides the estimation results. This output result provides useful
data on estimation of fertilizers cost/quantity, water, seeds, cultivation equipment cost
and Labour Day efforts/cost with Labour Day effort distribution on calendar chart of
crop life cycle, crop yield along with extrapolated market price at the harvest time and
its profitability. All the required inputs which are both linear and non-linear in nature
are taken by farmer’s data base, external information sources mentioned earlier. The
inputs get processed by machine learning techniques and generate the estimation with
feasibility study so that the farmer can choose the desired crop for cultivation.

 Crop care service: The crop care service guidance spans right from the sowing of
seeds as start point till the time of harvesting as endpoint. The complex structured
data sampled from IoT sensors from the fields are analyzed along with the data
collected from sources of information sites along with domain expert inputs wherever
needed through Artificial Intelligence techniques. After the analysis of complete data,
the overall corrective action item is derived out of PID (Proportional Integral &
 Differential) controller mechanism. Accordingly, the corrective measures are alerted
to the farmer on their smart phone to prioritize the action based on severity and
urgency to act upon.

 Price prediction and market guidance: This service helps to safeguard the farmers
from market fluctuation and mitigates the risk of price loss. Based on the statistical
data collected from various sources a predictive price and demand information is
shared with the famers during the complete crop lifecycle. And hence the farmers can
plan better for releasing their commodities to market.

 Crop loan and insurance service: This service helps the farmers in facilitating
feasibility of the getting, crop loan, processing support, eligibility criteria and loan
limit as per the smart estimation made for the propose crop. Also, it helps to get the
crop insured as a mitigation plan for crop failures due to any uncertainties or
calamities.
Currently used AI technologies in Agriculture

 Blue River technology: Founded in 2011. This California-based startup combines

artificial intelligence, computer vision and robotics to build next-generation
agriculture equipment that reduces chemicals and saves costs. Computer vision
identifies each individual plant, decides how to treat each individual plant and
robotics enables the smart machines to take action. The use of sensors that detect
weeds, the type of weeds and the right herbicides to apply within the right buffer
around the plant. The cameras and sensors use machine learning where the images are
captured and the machines can be taught in different weeds. Then also the right
herbicides are sprayed precisely as per encroachment area. Blue River Technology
has developed a robot called See & Spray which reportedly leverages computer vision
to monitor and precisely spray weeds on cotton plants. Precision spraying can help
prevent herbicide resistance. The short video below demonstrates how the robot
works in action.

 FarmBot: Founded in 2011. This company has taken precision farming to a different
level by enabling environment conscious people with precision farming technology to
grow crops at their own place. The product, FarmBot comes at a price of $4000 and
helps the owner to do end-to-end farming all by himself. Ranging from seed
plantation to weed detection and soil testing to watering of plants, everything is taken
care of by this physical bot using an open source software system.

 Harvest CROO Robotics – Crop Harvesting: Harvest CROO Robotics has

developed a robot to help strawberry farmers pick and pack their crops. Lack of
labourers has reportedly led to millions of dollars of revenue losses in key farming
regions such as California and Arizona. The robot picks up strawberries, helping
farmers reduce the cost of harvest labour. Strawberries need to be picked in a certain
time period and hence qualified pickers are needed. Harvests CROO Robotics
believes that their invention will save money, increase yields, reduce energy usage
and improve quality. Watch this short vision and learn more.
  Plant diseases diagnosis app - Plantix: The Berlin-based agricultural tech
startup PEAT developed the Plantix app that identifies potential defects and nutrient
deficiencies in soil. The app uses images to detect plant diseases, a smart phone
collects image which is matched with a server image and then a diagnosis of the plant
health is provided. In this way the application uses AI and machine learning to solve
the plant diseases.

 Prospera: Founded in 2014. This Israeli startup has revolutionized the way farming is
done. It has developed a cloud-based solution that aggregates all existing data that
farmers have like soil/water sensors, aerial images and so on. It then combines it with
an in-field device that makes sense of it all. The Prospera device which can be used in
green houses or in the field, is powered by a variety of sensors and technologies like
computer vision. The inputs from these sensors are used to find a correlation between
different data labels and make predictions.

Challenges with Artificial Intelligence in Agriculture

Though Artificial Intelligence offers vast opportunities for application in agriculture,

there still exists a lack of familiarity with high tech machine learning solutions in farms
across most parts of the world. Exposure of farming to external factors like weather
conditions, soil conditions and presence of pests is quite a lot. So what might look like a good
solution while planning during the start of harvesting may not be an optimal one because of
changes in external parameters?
AI systems also need a lot of data to train machines and to make precise predictions. In
case of vast agricultural land, though spatial data can be gathered easily, temporal data is hard
to get. For example, most of the crop-specific data can be obtained only once in a year when
the crops are growing. Since the data infrastructure takes time to mature, it requires a
significant amount of time to build a robust machine learning model. This is one reason why
AI sees a lot of use in agronomic products such as seeds, fertilizer, pesticides and so on rather
than in-field precision solutions.
Conclusion

AI technologies help farmers to analyse land/soil/health of crop etc and save time and
allow farmers to grow right crop in each season that has best yield. Vertical cropping can
reduce water usage, make efficient land usage, can be cultivated in urban areas in buildings.
It can reduce the problems with labour unavailability. Allows prediction of next year crop
seasons/weather/climate/rainfall etc. AI based predictions enable suggesting appropriate
pesticides/crops/place at right time before large scale incidence of disease. With a huge space
still untouched in agriculture for the intrusion of automatic response systems, there is a vast
opportunity for the agriculture industry to leverage emerging technology of catboats for
assisting farmers with the answers to all their queries and giving relevant advice and
recommendations to their specific farm related problems. This in turn propels the growth of
the AI market in agriculture.