See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312589560

Status and Scope of Robotics in Agriculture

Conference Paper · December 2016

CITATIONS READS

8 10,401

7 authors, including:

H. L. Kushwaha Jagnnath Prasad Sinha

Indian Agricultural Research Institute Indian Agricultural Research Institute
20 PUBLICATIONS   157 CITATIONS    57 PUBLICATIONS   134 CITATIONS   

SEE PROFILE SEE PROFILE

Tapan Khura Dilip Kumar Kushwaha

Central Agricultural University Indian Agricultural Research Institute
18 PUBLICATIONS   265 CITATIONS    11 PUBLICATIONS   20 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of Drone (UAV) Assisted Sprayer View project

Development of solar powered pneumatic seed cleaning system View project

All content following this page was uploaded by Jagnnath Prasad Sinha on 22 January 2017.

The user has requested enhancement of the downloaded file.

 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

29

Status and Scope of Robotics in Agriculture

H. L. Kushwaha, J. P. Sinha, T. K. Khura, Dilip Kumar Kushwaha, Uttpal Ekka,
Mayank Purushottam and Nishtha Singh

Division of Agricultural Engineering, ICAR - Indian Agricultural Research Institute, New Delhi-
110012,
E-mail: hlkushwaha@gmail.com

ABSTRACT

Robotics is playing a significant role in agricultural production and management. There is a

need for autonomous and time saving technology in agriculture to have efficient farm
management. The researchers are now focusing towards different farming operational
parameters to design autonomous agricultural vehicles as the conventional farm machineries
are crop and topological dependent. Till date the agricultural robots have been researched and
developed principally for harvesting, chemical spraying, picking fruits and monitoring of
crops. Robots like these are perfect substitute for manpower to a great extent as they deploy
unmanned sensing and machinery systems. The prime benefits of development of
autonomous and intelligent agricultural robots are to improve repeatable precision, efficacy,
reliability and minimization of soil compaction and drudgery. The robots have potential for
multitasking, sensory acuity, operational consistency as well assuitability to odd operating
conditions. The study on agricultural robotic system had been done using model structure
design mingled with different precision farming machineries. Few prototypes were designed
by European Union named CROPS, USA-ISAAC2 & Michigan-Hortibot, Australia-AgBot,
Finland- Demeter, India-Agribot and many other countries. The agricultural robots are
designed using different localization techniques which are vision, GPS, laser and sensor
based navigation control system. In this paper, comparative study including an overview of
Robotics approach for precision Agriculture in India and worldwide development is explored.

Keywords: Precision Farming, Agricultural Robots, localization techniques, Mechatronics.

1. INTRODUCTION
Automation of agricultural operations is demand of the time to improve the productivity with
the help of tools and technology. In recent years, the development of autonomous vehicles in
agriculture has experienced increased interest. Many researchers started developing more
rational and adaptable vehicles for agricultural operations. In the field of agricultural
autonomous vehicles, a concept was adopted to use multiple small efficient autonomous
machines in place of traditional large tractors [9]. Moreover,such asystem may have a less
environmental impact as it can reduce over-application of chemicals and high usages of

264
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

energy and inputs by the control that is better matched to stochastic requirements. There are
numbers of field operations that can be executed by autonomous vehicles, giving more
benefits than conventional machines. Most of the researcher are working for autonomous
vehicle design for precision agricultural mobile robots [9][11][12] . The design works on
implementing three different verticals namely 1. Mobile robot navigation 2. Implements
(Framework & Applications) 3.sensor modules. Different countries like the USA, European
Unions, Denmark, Australia, Finland, India etc are designing mobile robots under these
verticals which are mainly to procure agriculture farming over commercial industries.
Research groups have developed differentspecialized navigation techniques like an odometer,
vision based,sensor based, inertial, active beacon, GPS, map-based, landmark navigation
techniques to operate robots under unified controlspace for farming. This technique is used
for application likeseed-bed preparation,seed mapping,seed placement, reseeding,
cropscouting, weed mapping, robotic weeding control, micro-spraying, robotics gantry,
robotic irrigation, etc. [10]
In thescientific literature,studiesshow that to adapt agricultural machinery for agricultural
platforms (autonomous vehicles or mobile robots) [30] [21] [26] leads to the development of
autonomous navigation for agricultural machine should be designed with highersafety. In
recent trend in the development of platforms specifically built for agricultural autonomous
vehicles or robots [3][4]. Autonomous robot research in agriculture is well represented in the
literature. Excellent research carried out in Canada, Japan, Europe, Australia, the USA and
initiated in India development of agricultural autonomous vehicles [13][14] [21][9][1]. Most
agricultural autonomous robotics research has been performed in controlled
environmentssuch as robotic picking of cherry tomatoes [23], cucumbers [27], mushrooms
[31] and other fruits [23]. In horticulture, robots have been applied to citrus [21] and apples
[31]. Also, milking robots have had much attention, particularly in the Netherlands [32].
However, the development of these platforms presents two challenges [9]: to develop a
physical structure suitable for the agricultural environment, and to develop an electronic
architecture to integrate the various electronic devices. An electronic architecture must be
robust and reliable, provide quick and ease maintenance and have modularity and flexibility
to allow future expansions and connection of new equipment. As more autonomous
applications for agriculture will be operating in upcoming years, it gives an idea applying
robotics in agriculture may be technically difficult compared to industrial robots.
This paper provides information about the currentstatus of autonomous agricultural robots
designed in different countries as well as in India. Alsoscope of new trends of designed
robots for farming in agricultural sector.
1.1. Curent sceanarios of Agriculture system in India and Factors Affecting
The projected world's population to grow to more than 9.15 billion by 2050 [25]. Therefore,
the challenge for the next decades will be tosupply the needs of the expanding world
population by developing a highly productive agriculture management, whilst at the same
time preserving the quality of the environment [37]. Most of the developing countries
including India facing agricultural labor shortage problem. A major portion of youths from
village shifting to urban for led better life. As a result, agriculture operation gets delayed
during its peakseasons due to a labor shortage. Human, animal, and Mechanical power source
are utilized in agricultural operation in terms ofseed bed preparation, tillage,seeding or
transplanting, fertilizer and chemical application, intercultural operation and harvesting.
Average power availability on Indian farm is about 2.02 kW/ha. Human, animal, and

265
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

Mechanical power operated agricultural machinery fail to optimize the operational cost, time
and all other inputs. The tractor is the main machine powersource on which most of the farm
mechanization depends. Most of the farm operation is done by using the tractor as a prime
mower with attaching different implements on it. Although the traditional animal operated
country plow gives low output and requires a higher number of field operations are still being
used by the majority of the farmers. Zero tillseed drill, inclined plate planter is used
forseeding operation which is tractor power operated. High precision and input required
during intercultural and plant protection operation. Placement ofseed at optimum depth for
better germination rate required precession applicator. At present intercultural operation
mainly carried out by using manually operated implements such as Wheel hoe and
Conoweeder are very popular for horticultural crops and paddy field. Both the implements
are manually drawn. On the other hand, nutrient management is the important operation for
augmenting agricultural production and retain itssustainability. Most of the farmers in
developing country apply nutrients manually withoutsimulating the already availability of the
nutrient in the field and application of nutrient is uncontrolled throwing by the hand.some
instrumentalsetup has been designed for delivering fertilizer which facilitates for delivering
fertilizer uniformly.
2. CURRENT STATUS OF AGRICULTURAL ROBOTS
2.1. MF-Scamp Robots Designed by Blackmore
MF-Scamp robots are designed for scouting, weeding and harvesting [28]. It is designed
either four wheel or sixwheel drive weed seeking robot to perform weed removing or destroy
the weed. Generally the crops are grown inrows and column which can remove weeds when
running hoe tools between the crops row and column. Now this intelligent hoe tools uses
vision sensor to locate and identify the crops in rows and column and steers itself accurately,
to a larger extend reducing the usage of herbicides. It also uses colorsensor to identity weeds
between the crops by producing weed maps identifying plans. This robot designed not only
reduce the labor time but also the economic feasible withslight reductions in prices of
navigation systems. This kind of design not only reduce the cost of spraying but also decrease
the usage of tractor. The negative effects of the robot is the higher costs for small farmers and
additional costs used for electronics devices like GPS-system as shown in Fig. 1.

Fig. 1. MF-Scamp Robots Designed by Blackmore [27]. Fig. 2. API Platform [27].

266
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

2.2. Autonomous Plant Inspection (API) Research Platform designed by Danish

institute of Agriculturalscience (DIAS)
The API Platform was initially developed by Madsen and Jakobsen in the year 2001. Further
it was developed by Aalborg University in Denmark. It is third generation autonomous
research platform for precision farming prototype designed having four wheel- drive, four-
wheel steering with two motors per wheel, one for providing traction and the other provides
steering to obtain high mobility in the movement of the robot [8] shown in Fig. 2. The robot
has 60cm height clearance, and a track width of 1m. It is also equipped with Real Time
Kinematic Global Positioning system (RTK-GPS) and there is an operating unit over the head
of the frame which implement for agricultural operations like spraying devices,sensors or
weeding tools. The navigation system is designed using farm management PC to compute
route plan in the field as well providing anti-collision system for the robot [4]. Based on
shape and size orientation of weeds it is possible to create weed patches of the entire field.
The weed patches are registered by using GPS and GIS systems to create weed maps of the
individual fields. The API platform has a speed of 3.6km/hr and capacity of 4.32ha/hr when
used in autonomous field scouting. This in turn enables weed mapping system to produce
weed maps for the robot and will be helpful for precise patch spraying with right mixture of
herbicides. The advantage of this kind of system will not only reduce the usage of herbicides
which might vary from 30% to 75%. However two issues occur one is to collect data about
weeds and another is to remove it physically. In farm land , there are different working space
one between crop rows known as inter-row area, another between the plants within the row
known as intra-row area and the close to crop area within leaf and root envelope. It is easy for
machine to operate at inter-row area as it is an openspace in line in the direction of robot
movement but intra-row area is very difficult to manage as thespace is delineated by
irregularspacing of crop plants. The machine doesn’t operate with increased care and
accuracy so that the crop plant tissue should not be damaged. Also the area close to
cropshould not have soil disturbance as this would lead to soil witting. The system should be
having intelligent sensor to implement microspraying or laser weeding. Laser weeding highly
used in mobile focused beam of infra-red light to thermally disrupt cell membranes in the
weed leaves and shoots.
2.3. Sub canopy robot ISAAC 2 from Hohenheim University, Germany
This prototype is designed to collect timely and accurate information in the crop carrying
range of sensors to assess crop health and status. This high clearance platform carries
instruments above the crop canopy and utilize GPS. The portal robot shown in Fig. 3 and 4
used to provide automated cropsurvey [4] as well as to measure crop nutrient status and
multispectral responses (stress), visible images (pan Chromatic), weedspecies and weed
density. Using machine vision the position and density (biomass) of different weed species
are recorded in the form of weed mapping. Two different methods are used to identity weeds
round the crop, first is to record the increased leaf area around weedy area as crops are
planted in rows [29] and secondly using active shape recognition, developed to identify
human faces, similarly can be used to identity weedspecies by the shape of their outline.
Currently 19 species of weeds can be recognized using color segmentation [36]. The final
result is the weed map created with the help of the machine vision and the robot knowing the
position of the weeds which can be removed, kill or retard from the unwanted area [27].
Different physical interaction techniques are used to remove the weeds one example is to
break the soil and root interface by tillage and promote wilting of weed plants. This is

267
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

achieved in the inter row area by using classical spring or duck foot tines where as for Intra
row weeding its more difficult as it requires the position of the crop plant to be known so that
the end effector can besteered away. Within the close-to-crop area, tillage cannot be used as
any disturbance to the soil is likely to damage the interface between the crop and the soil.
Different non-contact methods are being developed such as laser treatments [20] and micro-
spraying butstill completely removing of weeds is not achieved.

Fig. 3. Cropscouting platform Madsen and Jakobsen Fig. 4.sub canopy robot ISAAC2 [10]

2.4. BoniRob farming robot developed by Deepfield robotics funded by Bosch,

Germany
BoniRob is a multi-purpose robotic platform for applications in agriculture. It has four
independently steerable drive wheels that has the ability to adjust its track width and makes it
highly maneuverable as shown in Fig. 5 [6]. The robot run purely on batteries or connected to
a generator to extend its range and usage time. It can be retrofitted and upgraded with
exchangeable application modules (tools)[7]. The robot can navigate autonomously along
plant rows (e.g. dams) in the field, carrying the application module (plantation) as it goes. It
has environmental sensors like Lidar, inertialsensors, wheel odometry,satellite and GPS for
row detection and navigation. This robot is designed for application like chemically weed
control or uses a rod device to crush the weeds, to combat volunteer potatoes from the
previous year, to measure soil compaction and manage plant breeding. Testing of the robot in
carrot fields have proved that it is over 90% effective in removing weeds and preserving crop
plants. Communication devices for the robot are in development, which would make a
cohesive fleet of BoniRobs possible which will be able to communicate with each other as
well as human farmers[18]. Currently due to high costing of the design and sensor mounted
to the robot, the robot is used only for laboratory purpose and not yet commercialized.

268
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

Fig. 5. BoniRob (V2) during a RemoteFarming Fig. 6. Lettuce Bot attached to a tractor,
California
2.5. Lettuce Bot, California
Lettuce Bot also known as Rambo of weeds was designed by Heraud and the Bot’s co-
creator, Lee Redden atstandford University as shown in fig.2.6. The bot design is more of
robotics, computer vision and machine learning algorithm to advance the growing fields. The
Bot has a database of more than a million images that it uses to identify the plants. It looks
for a weed or a lettuce plant that are grown too close to plant or crops. Once it identifies the
weed, it releases a strong spray of fertilizer disintegrating the weeds around the plant and
fertilizing the lettuce around it[39]. This bot application can also be used in organic farming,
one of the areas where weeds are the biggest problem.since there’s no practical way of
eliminating weeds in an organic field other than pulling them by hand one by one which is
not sustainable and also fertilizer used are not organicso not allowed in organic fields. The
bot is designed for lettuce fields not for other fields as lettuce is the second largest crop
grown in California.
2.6. CROPS, European union
The CROPS research project, “Clever Robots for Crops”, which is sponsored by the EU
Commission, could provide a solution for an automated harvesting procedure. Other partners
are the Technical University of Munich, Wageningen University, the research institute, CSIC,
and the company FORCE-A [5]. The aim is a configurable, modular and intelligent robot
platform, which reliably recognizes both the fruit as well as obstacles and other objects. In
this way it can navigate and harvest on its own on plantations and in greenhouses. The robot
platform is also intended to be able to treat selected areas with pesticides. For the
development of the right gripping and cutting device, the researchers are concentrating on the
harvesting of apples and peppers [19]. They have developed two prototypes which uses an
adaptive gripping technology to adapt to the fruit varying insize and shape and recognizes
ripe fruit with the help of camera andsensors. The CROPS robotic platform is used for
selective harvesting of fruits which detects the fruit, determines its ripeness, then moves
towards the ripen fruit, grasps it and detach itsoftly, also it is capable of
determiningsitespecificspraying for targets spray towards foliage as shown in Fig. 7. It is also
design to develop techniques to detect and classify obstacles and other objects to enable
autonomous navigation and operation in plantations. Few challenges which CROPS

269
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

prototype faces is that it get difficult to identify the fruit hidden under leaves and also
grabbing and separating from ripen and unripens fruits due to their different shapes and sizes.

Fig. 7. Image processing design, CROPS. Fig. 8. HortiBot, Denmark.

2.7. HortiBot, Denmark

Hortibot is a commercially produced and robust tool carrier designed for high tech plant
nursing for e.g. organic grown vegetables as shown in fig.2.8. The steering of the bot is
operated using computer vision based guidance system. The main function is to control weed,
enable an automatic execution of one-sided repetitive weeding for outdoor gardening [16].
The HortiBot navigate on the basis of computer vision recognition of the topography of the
tracks between beds. Also it is capable of passing several parcels with visible rows based on
row detection system designed by Agrocom Vision with minimal use of GPS. The major
advantage of this prototype is that it did not take much time for labor training and reduces the
labor cost, doing repetitive tasks like cultivating and other form of mechanical weeding also
efficient use of herbicides thus reducing herbicides cost. But HortiBot also have some
drawbacks like highly cost ofshort term use, need professional operator to operate the
machines and can be used in small farms.still HortiBot is a new invention technology that
will be useful to increase the farming productivity.
2.8. AgBot II, Austrila
AgBot II is a robot designed to help farmers to take decisions on the use of herbicides,
pesticides, fertilizers and watering. It is developed at the Queensland University of
Technology (QUT) in Australia, usingsensor networks, drones, weather,satellite and
historical data to help “farmers” run mathematical models and statistical programs to help
and guide them in farm management decisionssuch as whether to use herbicides, pesticides,
fertilizers and how much water plants should be given as shown in Fig. 9. The bot is still a
prototype equipped with sensors and operated using software and can work individually or in
groups. Vision sensors, for example, allow robots to “see” whether a plant is a weed or a rose
and chemically or mechanically take action insupport of or against the plan.

270
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

Fig. 9. AgBot II, Australia Fig. 10. Vitiroversolar Robots, New Zealand.

2.9. Agribot, India

Many countries in Asia including India are agrarian economies and most of their rural
populations depend on agriculture to earn their livelihood. Agribot is an agricultural robot
designed by BIT Hyderabad students, India[1]. It is designed to increase the productivity,
speed, application accuracy of the work and minimizing the labor of farmers. Its major area
of function involved in farming i.e. harvesting,spraying,seeding and removing the weeds.
This robot is designed to execute the basic functions required to be carried out in farms.The
major design of the robot is based on image processing and relay controls the motor of the
robot. A vision based row-guidance method is used to guide robot platform driven along the
crops planted in row. The system is having a camera equipped to give a live vision of the
field, also for large farms a GPS based module can be installed depending on the map
specific land can be fixed for harvesting. After harvesting, it can pick up crops and place it in
a vessel, also spraying system can be designed on the robot which will spray the pesticides on
the crops. The main obstacle to this kind of robot farm is that farms are a part of nature and
nature is not uniform.
2.10. Vitirover solar Robots, New Zealand
A French company designed a smart autonomous robot called Vitirover as shown in fig.2.10.
This little autonomous robot uses thesolar power for the electrical motors which could work
for hours without any pause and are used for cutting grass & weeds in vines within 2-to-3 cm
of vine and has a speed of 500m/hr. For farmers, it is very important to use technology
friendly devices in the grape vines. Vitirover uses sensors and GPS system that keeps the
robot away from grape vines.
3. SCOPE OF AUTONOMOUS FARMING
3.1. Datascouting at different stage of crop growth
Yield in a given field may vary inspace depending on a combination of factors such as
nutrient availability, soil moisture, rooting depth, pest pressure, weed density, crop maturity
and others. Good agricultural practice needs an application of optimum input at appropriate
timeseries. Continue monitoring and data collecting related to crop NDVI, Biomass, Leaf
area index, crop growth rate, water stress are an important parameter for optimizing the
variable input parameters in different stages of crop growth and also crop health. Crop

271
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

physicalstatus monitoring would be less expensive and timelier if an automated system could
remain within the crop canopy for continual monitoring [1]. Real-time management of inputs
like fertilizers, herbicides, at variable rate could be achieved bysensing the different
parameters of the crop measured in the field by and processing it for further application part
using microprocessor or microcontroller. As the robotic vehicle continuously monitor the
crop canopy it is easy to identify crop diseases and pest attack at an early stage in patches or
selectively. Also, it can monitor weed density, waterstress at a different stage of crop growing
periods.
3.2. Intercultural operation
Weed competes with the crop for sunlight,space, and nutrients. To control weed species, a
large number of herbicides and chemicals are used in agricultural fields, which results in
drinking water contaminated and environmental pollution. Currently, the excessive use of
herbicides damages the health of people, animal etc. According to the weed science
research,about 33% of the total losses by agricultural pests are caused by weeds only[2].
Therefore it is important to identify the weeds from the crop andselectivelyspray herbicides
to optimize the chemical application. There are various techniques developed all over the
word for distinguishing weed from the crop.In the conventional weed control system, the
herbicide issprayed uniformly over the field which may damage crop condition. Machine
vision systems present a great potential to be used on data collection for precision agriculture,
where images would be used to extract information. The autonomous agricultural robot can
accomplish intercultural operation using the machine vision based technique. It can be
achieved by mapping the weeds inselectively or in patches by image processing method, in
row crop, the patch identification in between the row can help to distinguish the weed. There
is specificshape reorganization method are applicable for classifying the weeds by its shape
this method is more accurate in weed identification method. Another method is color
segmentation for identifying weeds. Based on all these technique weed maps is generated.
Detecting weeds selectively from a field with the application of proper methods of image
segmentation technique and post processing the segmented data finally, optimize the
application part of chemicals in the field.Based on the map generated appropriate weeding
technique is considered for weeding. There are different operational methods for removing
the weeds like mechanical chemical application. Identifying the actual position of the weeds
mechanical weeder can be incorporated in removing weeds by partial tillage operation at 2 to
4 cm depth. Removing of weeds from between the row is easy as compared to removing
weeds from intra-row which required high- speedsensing device and high- speed
mechanisms to push rotary blades or chemical spray for intra row application.
3.3. Fertilizer
Crop nutrient is a basic and important component ofsunlight, water for its growth and
increasing productivity. The rational use of fertilizers and fertilizer technology has a
significant effect on the development of agricultural [40]. Among the various nutrients, soil
nitrogen is one of the important parts of agricultural production, Nitrogen has been perceived
as a critical nutrient for productivity of crops but it also has important factors of
environmental contamination. A major portion of the input fertilizer is lost by leaching, and
contaminates freshwater, marine ecosystems when high rates of N fertilizers are applied to
agricultural fields [37].There are variations of nutrient in quantity within the small area of the
field. Traditional measurements of soil nitrogen are very difficult and laboratory method by
which nutrient map is generated. Which helps in pre-assessment of the actual requirements of

272
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

fertilizer in the field. That involve in the uniform rate of application of fertilizer over an
entire field. This map used as asite-specific management tool that allows the producer to vary
the rate of application of crop input across a given area. Variable-rate technology has a
potential to optimize the input that further enhance the benefits of operator and maintain soil
environmental health. Most of the Variable rate technology are GPS base few are available
on control system based on reckoning up track distance for a variable rate fertilizer applicator
without GPS. Nitrogen variability in an area is defined by grid area mapped by GPS latitude
and longitude. According to a position of pre-defined grid area nitrogen output is fixed in
controlling unit. While during the time of final fertilizer application controlling unit have
control over the motor rpm which directly responsible for opening are of metering unit or its
rpm that is calibrated with the input amount of fertilizer requirement at that particular grid.
3.4. Harvesting
Bulk harvesting is the common trend of harvesting in India. This process of harvesting
suitable for few selective crops,so farsome crops like cotton where the maturity of cotton ball
achieves at the different time period in asingle plant [33] therefore in this case selective
harvesting of a cotton ball is important. Other than the cotton cereal pulses crop are get
matured in patches selectively. Harvesting immature crop effects its quality. Therefore
selection of mature crop for harvesting needs certain threshold parameter to define the range
of maturity level for defining crop as mature.selective harvesting needssensing technology
for collecting information of crop which needs to be processed through microprocessor or
microcontroller and finally define thestatus according to which decision support system
supply command to the mechanical mechanism to harvest of defined crops. Broadly there is
two part of work first is tosense the cropstatus and according to that harvesting of aselective
crop is done which is thesecond part of this complete process. Harvesting with the
agricultural robot can be done at real-time datasensing and the processing unit or feeding the
processed data map on robotic vehicle harvesting could be done.such kind of precision
monitoring and harvesting with the help of agri-robot could help to get ride off the ever
decreasing labor scarcity load and increasing labor charge and also can maintain a quality of
food at the end of the harvesting.
Fig. 11 gives a brief idea of the Architecture of Agricultural Robots. The Bots uses GPS,
Vision processes, Lidar etc. for navigation in the field. It calculates the waypoints of the field
and accordingly robots move across and perform the given operation. The basic principle for
the agricultural robot is thestability factor while moving in the field are controlled by the
speed & rpm of the motor, obstacle detection as well controlling the angular and linear
movement of the robot. The main source for power is fromsolar and battery. This factors
leads to field operation like optimum seed placement, weed control,spraying, harvesting, crop
monitoring etc.

273
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

Fig. 11. Architecture of Agricultural Robots.

4. CONCLUSIONS
The autonomous robot have potential to work on precision agriculture having continuous
monitoring by using different sensing technology, which provides different crop status
parameter like, micro nutrient availability, bio mass index,status of pest and disease,
waterstress, thermalstress etc. for better remedies of crops. As ever increasing of world
population and decreasing of agricultural workers created constraint to farming system.
Agricultural robot has potential to take off the load of labor shortage and increasing the
productivity. Various technology like machine vision, image processing, and mechatronics
can be assembled in asingle platform named agricultural robot which can give optimal
solution for autonomous agricultural operations.

REFERENCES
[1] Akhila Gollakota, M.B.srinivas, 2011.AgriBot- A Multipurpose Agricultural Robot.
Annual IEEE India conference,Hyderabad.
[2] Amruta A. Aware, Kavita Joshi, 2016. Crop and Weed Detection Based on Texture
andsize Features and Automaticspraying of Herbicides. International Journal of
Advanced Research in Computerscience andsoftware Engineering.

274
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

[3] Åstrand, B.and A.J. Baerveldt 2002. An agricultural mobile robot with vision-based
perception for mechanical weed control. Autonomous Robots 13(1):21-35.
[4] Bak, T. and Jakobsen, H. 2004. Agricultural Robotic Platform with Four Wheelsteering
for Weed Detection. Biosystems Engineering 87(2) 125-136.
[5] Bakker T., 2009. An Autonomous Robot for Weed Control. Wageningen-The
Netherlands Wageningen University.
[6] Bangert W., A. Kielhorn, Dr. F. Rahe, F.sellmann, W.strothmann, Prof. Dr. D. Trautz,
2010. Field-Robot-Based Agriculture: “RemoteFarming.1” and “BoniRob-Apps”,
Bosch, University of Applied
[7] Sciences Osnabrueck, Osnabrueck.
[8] Billingsley, J., Visala, A., Dunn, A., 2008. Robotics in Agriculture and Forestry, in
“Springer Handbook of Robotics”, pp. 1065-1078 (Editorssiciliano, B., Khatib,
O.),springer.
[9] Bisgaard, M., Vinther, D. and Østergaard, K.Z. 2004. Modelling and Fault-Tolerant
Control of an Autonomous Wheeled Robot. Group Report 04gr1030a, Institute of
Control Engineering, Aalborg University, Denmark.
[10] Blackmore, B.s., Fountas S. , Vougioukas S. , Tang, L.,sørensen, C. G., and Jørgensen,
R. 2004b, Decomposition of agricultural tasks into robotic behaviors. The CIGR
Journal of AEscientific Research and Development in Press.
[11] Blackmoresimon, Billstout, Maohua Wang, Boris Runov, June,2005. Robotics
Agriculture- The Future.
[12] of Agricultural Mechanisation? 5th European Conference on Precision Agriculture
Uppsala,sweden 9-12th.
[13] Blackmore S. , Griepentrog, H.W., Fountas S. , Gemtos, T.A., 2007.specifications for
an autonomous crop mechanizationsystem. Agricultural Engineering International: the
CIGR Ejournal. Manuscript PM 06 032. Vol. IX.september.
[14] Burks, T. F.,shearer S. A., Heath, J. R. and Donohue, K. D. 2005. Evaluation of
Neuralnetwork Classifiers for Weedspecies Discrimination. Biosystgems Engineering
91(3), 293-304.
[15] Cembali, T., J.R., Folwell, T. Ball and D.D. Clary. 2005. Economic Comparison of
selective and Non-selective Mechanical Harvesting of Asparagus. ASAE Paper No.
053003.st. Joseph Mich.: ASAE.
[16] Chi, Y. T. and P. P. Ling. 2004. Fast Fruit identification for Robotic Tomato Picker.
ASAE Paper No: 043083.st. Joseph Mich.: ASAE.
[17] Christensen, L. K. S. K. Upadhyaya, B. Jahn, D. C.slaughter, E. Tan, and D. Hills.
2005. Determining the influence of water deficiency on NPKstress discrimination in
maize usingspectral andspatial information. Precision Agriculture Journal 6: 539550.
[18] Claus G.sørensen, Michael Nørremark, RasmusNyholmJørgensen, Kjeld Jensen,
2007.Hortibot: Feasibility study of a plant nursing robot performing weeding operations
– part IV. ASABE Annual International Meeting.
[19] Godwin, R. J., Earl, R., Taylor, J. C., Wood, G. A., Bradley, R. I., Welsh, J. P.,
Richards, T., Blackmore, B.s., Carver, M. C., Knight S. M., and Welti, B. 2001.
Precision Farming of cereal crops: A five-year experiment to develop management

275
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

guidelines. Project report 264e, -328pp, London, Home Grown

CerealsAuthority.http://www.hgca.com/publications/documents/cropresearch/267_Com
plete_Final_Report. pdf.
[20] Griepentrog, H.-W., Ruckelshausen, A., Jörgensen, R.N., Lund, I., 2010.
Autonomoussystems for plant protection, in „Precision Crop Protection – The
Challenge and Use of Heterogeneity“, pp. 323-334 (Editors Oerke,E.-C., Gerhards, R.,
Menz, G.,sikora, R.A.),springer.
[21] Heisel, T., Christensen S. , Walter, A. M. 1999. Whole-field experiments withsite-
specific weed management. In: ECPA 02: Proceedings of thesecond European
Conference on Precision Agriculture, Odense, Denmark, 11–16 July, 1999 (stafford J
V, ed), Part 2, pp 759–768.
[22] Heisel, T., 2001, Weeds insugar beet rows - I. Influence of neighbor plant on the beet
yield - II. Investigations of a CO2 laser for in-row weed control, 56, Denmark, DIAS,
DIAS Report Plant Production.
[23] Hannan M.W., T.F.Burks., 2004. Current Development in automated citrus harvesting,
ASAE Paper No:04-3087,st. Joseph, Mich, ASAE.
[24] Keicher R., H.seufert, 2000. Automatic guidance for agricultural vehicles in Europe.
Computers and Electronics in Agriculture 25,169–194,Elsevier.
[25] Kondo N., Y. Nishitsuji, P. P. Ling, K. C. Ting, 1996. Visual Feedback Guided Robotic
Cherry Tomato Harvesting, the Americansociety of Agricultural and Biological
Engineers,st. Joseph, Michigan.
[26] Madsen, T. E. and Jakobsen, H. L. 2001, Mobile Robot for Weeding, Unpublished
MSc. thesis Danish Technical University.
[27] Murakami N., A. Ito, Jeffrey D. Will, Michaelsteffenb, K. Inoue, K. Kitaa S.
Miyauraa, 2008. Development of a tele-operationsystem for agricultural vehicles,
computers and electronics in agriculture.81–88, Elsevier.
[28] Nikos Alexandratos and JelleBruinsma,2015.World agriculture towards 2030/2050: the
2012 revision.FAO Agricultural Development Economics Division.
[29] Nørremark, M. and Griepentrog, H. W., 2004. Physical methods to control weeds
within crop rows, AgEng04, Leuven, Belgium.
[30] Pedersens. M., Fountass. and Blackmores., 2008. Agricultural Robots-Applications and
Economic Perspectives,service Robot Applications. Yoshihiko Takahashi (Ed.), ISBN:
978-953-7619-00-8, InTech.
[31] Pedersen, B. B., 2001. Weed density estimation from digital images inspring barley,
Unpublished MSc thesis KVL, Denmark.
[32] Reid John F., Qin Zhang, Noboru Noguchi, Monte Dickson, 200. Agricultural
automatic guidance research in North America. Computers and Electronics in
Agriculture 25,155 – 167, Elsevier.
[33] Reid J.N. S. J. Miles, J.Butler, M.Baldwin,R. Noble, 2001. Automatic Mushroom
harvesting Development. Journal of Agricultural Engineering Research, 78:15-23.
[34] Rossing, W., P.H. Hogewerf, 1997.state of the art of automatic milkingsystems,
Computers and Electronics in Agriculture, Volume 17, Issue 1, pp 1-17.

276
 International Conference on Emerging Technologies in Agricultural and Food Engineering
27 – 30th December, 2016, Agricultural and Food Engineering Department, IIT Kharagpur

[35] Robert.B.hutmacher, Ron N. Vargas,steven D. Wright, Bruce A. Roberts, 2003. Harvest

aid materials.
[36] and practices for California cotton a study guide for agricultural consultants and pest
control advisers. University of California.
[37] Rubens Andre Tabile, Eduardo Paciencia Godoy, Robson Rogério Dutra Pereira,
GiovanaTripoloniTangerino, Arthur José Vieira Porto, Ricardo Yassushi. Inamasu,
2010. Design of the mechatronic architecture of an agricultural mobile robot,5th
IFACsymposium on Mechatronicsystems Marriott Boston Cambridge Cambridge, MA,
USA,sept 13-15.
[38] Søgaard, H. T. and Heisel, T., 2002. Weed classification by activeshape models. AgEng
2002, International Conference on Agricultural Engineering, Budapest, Hungary, June-
July 2002.
[39] Tang, L., Tian, L., andsteward, B. L., 2000. Color imagesegmentation with genetic
algorithm for in-field weedsensing , Transactions of the ASAE - Americansociety of
Agricultural Engineers 43:41019-1028.
[40] Tilman, 1999 . Global environmental impacts of agriculture expansion; the need
forsustainable and efficient practices. Proceedings of the National Academy ofsciences,
USA 96, 5995–6000.).
[41] Tim Dyson, 1999. World food trends and prospects to 2025. the National Academy
ofsciences colloquium ‘‘Plants and Population: Is There Time?’’ held December 5–6.
[42] Wheeler T.R., P. Hadley, R.H. Ellis, J.I.L. Morison, 1993. Changes in growth and
radiation use by lettuce crops in relation to temperature and ontogeny. Agricultural and
Forest Meteorology, 66,173-18.
[43] Yueling Zhao, Liying Cao, Guifen Chen, Haiyan Han, 2011. Precision
Fertilizersharingsystem Based on Oracle Database. International Conference on
Mechatronicscience, Electric Engineering and Computer, Jilin, China.

277

View publication stats