Artificial Intelligence in Agriculture 8 (2023) 46–59

Contents lists available at ScienceDirect

Artificial Intelligence in Agriculture

journal homepage: http://www.keaipublishing.com/en/journals/artificial-
intelligence-in-agriculture/

How artificial intelligence uses to achieve the agriculture sustainability:

Systematic review
Vilani Sachithra, L.D.C.S. Subhashini ⁎
University of Sri Jayewardenepura, Sri Lanka

a r t i c l e i n f o a b s t r a c t

Article history: The generation of food production that meets the rising demand for food and ecosystem security is a big
Received 3 February 2022 challenge. With the development of Artificial Intelligence (AI) models, there is a growing need to use them to
Received in revised form 20 April 2023 achieve sustainable agriculture. The continuous enhancement of AI in agriculture, researchers have proposed
Accepted 25 April 2023
many models in agriculture functions such as prediction,weed control, resource management, advance care of
Available online 28 April 2023
crops, and so on. This article evaluates on a systematic review of AI models in agriculture functions. It also reviews
Keywords:
how AI models are used in identified sustainable objectives. Through this extensive review, this paper discusses
AI considerations and limitations for building the next generation of sustainable agriculture using AI.
Agriculture © 2023 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co., Ltd. This is an open
Sustainability access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Review
Robotics
Deep learning

1. Introduction and mitigating atmospheric changes is the biggest global challenge in

the 21st century (Di Vaio et al., 2020). The terms “sustainability” and
The agricultural sector in any nation plays a key role to address the “sustainable” has gained substantial attention applied in varies contex-
one of universal challenges, provide sufficient foods to survive people. tual aspects (Bolis et al., 2014). Sustainability is defined as a balanced
As estimated (Alexandratos and Bruinsma, 2012), in 2050, there is a re- combination of social, environmental and economic performance to
quirement to increase global food supply by 60% in order to feed nearly benefit current and future generations (Geissdoerfer et al., 2017). To
9 billion people (Padilla and Hudson, 2019). Growing population leads safeguard food and ecological security, the sustaining of performing
continuous farming with limited arable land (Jayne et al., 2014). This more of the same thing is commonly indicated as sustainable agricul-
issue is further aligning with the 17 Sustainable Development Goals ture (Gaffney et al., 2019). Thus, to achieve sustainable growth in the ag-
(SDGs) which has been focused to eliminate poverty and eradicate hun- riculture sector has received greatest attention (Castro and Swart,
ger and malnutrition by 2030 and 2025 respectively. Growing popula- 2017), and there is an emerging consumer demand for sustainable qual-
tion leads continuous farming with limited arable land (Padilla and ity food products (Mangla et al., 2019). Sustainability lays in three pillars
Hudson, 2019). It has been argued that food production process creates economic, social and environmental performance. Social performance
a foremost universal environmental degradation created through fertil- focuses on social troubles namely human rights, ethics in doing
izer utilization, greenhouse gas emissions and biodiversity (Tilman business, environmental activities, identical opportunities concerns on
et al., 2011). Though intensive agriculture (known as intensive farming) waste generation, greenhouse gas emission; economic performance
and industrial agriculture have led to an increase in food production and quantifies operational efficiency, shareholder value and transaction
easing of food shortages, now bring disadvantages due to utilization of costs (Ala-Harja and Helo, 2015; Panda, 2014). Subsequently, Sayer
high input of fertilizers, pesticides and fresh water (Tian et al., 2021). and Cassman Sayer and Cassman (2013) opined that agricultural
In particular, climate changes such as global warming, aggravating firms/farms require to obtain four objectives, which are often to be
flooding and drought will in turn influence the food security (Wheeler competing each other, to be considered sustainable growth namely
and Von Braun, 2013). Consequently, how to feed the increasing popu- 1) Ensure production of an adequate food supply; 2) Alleviate poverty;
lation while decreasing the negative consequences on the environment 3) Achieve better health and nutrition for a growing population; and
4) Conserve natural resources. These objectives are highly relevant to
⁎ Corresponding author.
the sustainable pillars, and they are aligned properly with the SDGs.
E-mail addresses: vilani@sjp.ac.lk (V. Sachithra), subhashini@sjp.ac.lk To certify the sufficient food productions to the growing population,
(L.D.C.S. Subhashini). technologically advanced inputs, cultivation techniques and soil

https://doi.org/10.1016/j.aiia.2023.04.002
2589-7217/© 2023 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co., Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

management approaches become the vital sources (Gaffney et al., to identify the firsthand experiences on currently applied AI technolo-
2019). High-yielding cropping systems must be concerned to convert gies in agriculture sector and variety of AI technology initiatives to
resources to economic yield. Increasing protein obtainability of food achieve sustainability growth objectives in agriculture sector. To carry
sources of food (beans, vegetables, wheat, rice), and of availability of vi- out this SLR, we set up three stages namely, planning; implementing
tamins and minerals through poise diet should require to ensure and reporting (Ferreras-Fernández et al., 2013).
healthy and affluence over 9 billion people people to be fed by 2050. In the planning stage, described the key terms that could be
Soil degradation, low irrigation management, and a less productivity considered relate to the study namely, agriculture, farming, protected
sludge farmers in poverty (Tittonell and Giller, 2013). To eradicate pov- agriculture, smart farming, Artificial Intelligence (AI), Deep Learning
erty, agricultural sector must be move forwarded the modernized and (DL), Machine Learning (ML), agricultural robot and robotics. The Bool-
productive agricultural transition where farmers are equipped with ean operators AND and OR were used to do more thorough searches, for
highly resourceful and resilient. Conserving natural resources consists example, “AI” AND “agriculture” AND “crops” OR “farming” AND “Smart
of wide range of soil nutrient, quality of water, green-house gas, Agriculture” AND “smart farming”. The search was performed in the
confrontation of pest and weed and reduction of aquifer. Preserving bio- four well-known data sources that encompass multidisciplinary
diversity of natural settings, flora and fauna are conservation challenges. publications, google scholar, Scopus, Science Direct and Web of Science,
Further, when underpinning technologies are infantile or improperly following the process used by similar recent studies of AI and sustain-
used, agricultural expansion causes a serious environmental damage. able agriculture (Traldi, 2021; Navarro et al., 2020). The scope of the
This is why Mellor (Mellor, 2017) insisted that identifying and utilizing publications was limited to documents such as journal and conference
vigorous pattern of technology improvements and efforts should focus articles, published in English. The past ten years considered as the
not on just one aspect of sustainability objective, but rather on activa- time range to conform the objectives of the study.
tion of the whole system that representing the prevailing agricultural In the implementing phase, 1421 articles were selected with the
enterprises. Yet, many firms/farms in the agricultural sector are stressed search tool. A database review of publications about the desired key-
with squat profit and low productivity (Barth et al., 2021), hinders words in Web of Science found 347 records since 2012. In the case of
efforts to sustainable agriculture (McGuire, 2017). Gaffney et al. Scopus and Science Direct, there have been 256 and 244 documents
(Gaffney et al., 2019) further stressed that growth in emerging and published respectively. The total number of articles published in Google
recently emerged markets (Asia and Africa) creates the definitive re- scholar search engine was 574 from 2012 to 2021 December. The areas
straints to meet sustainability objectives. Regardless of the complexity in which they have been published the most are agriculture, technology,
to meet all four objectives simultaneously, agriculture sector is moving computer and electronic, agronomy, agriengineering, computer sci-
towards sustainable agriculture (Tian et al., 2021). Realizing and utiliz- ences and sustainability. After getting the articles, they were manually
ing technological advancements, the commitment derive to agriculture reviewed through the title, keywords, abstract and text analysis adher-
sustainability must be accompanied with technological improvement ent to the objectives proposed of the study. Number of record screened
(Mellor, 2017). Thus, to meet Sayer and Cassman's sustainable objec- was 313. During this process, the list of documents was consequently
tives and face the global food security challenges ahead, wider applica- sorted to eliminate the duplicate articles (Åstrand and Baerveldt,
tion of existing technologies and utilization of advanced technological 2002). 131 articles were excluded due to irrelevant to agriculture indus-
tools and techniques soon would be the straightforward strategies try. This analysis resulted in 115 articles deemed eligible which were
(Franco, 2021). What we observed from present panorama is that that incorporated as a sample for this study. Out of 115 articles, 45 identified
Covid-19 pandemic emerges as a great crisis, leading to widen global through Web of Science, 37 Science Direct and 33 identified through
food security issue. Although the generation of food production that other sources listed above. The article list was finalized in December
meet the rising demand for food and ecosystem security is a big 2021 (Refer Table 1 in Appendix 1).
challenge, rapid developments in technology are making it possible. Re- In the reporting stage, each of the articles retrieved was analyzed
searchers applied artificial intelligence (AI) to make sustainable agricul- according to the AI component such as DL, ML, neutral network and
ture (Li et al. 2021b; Mohapatra and Lenka, 2016). The recent robotics and agriculture activities namely harvesting, plant eco-
application of the technologies of AI support to provide solutions to phenotyping, grading system, weed and crop classification, disease
problems in agricultural domain. These technologies are used to reduce detection and monitoring and soil management. The searched articles
the cost as well as increase the effectiveness and efficiency. There are were then listed on excel spreadsheet. The data sheet contained the de-
surveys which conducted to find what people did to make sustainable tails of article namely name of author/s, year of published, study title,
agriculture using AI. However, investigations on how AI used to achieve key AI technology used, main agriculture area, benefits obtained and
sustainable objectives are still under research-able area. Specifically, limitations. Once the database was completed, a content analysis was
this research aims to map and create an understanding of the various performed to examine the review summary in-depth and summarise
technologies implement in agriculture sector with a special focus on the empirical experiences on currently applied AI technologies in agri-
the sustainability growth objectives. Thus, the main purposes of this culture sector and variety of AI technology initiatives to achieve sustain-
systematic review are are to; 1) develop a more complete understand- ability growth objectives in agriculture sector. Fig. 1 illustrates the
ing of the enabling AI technologies currently applied in agriculture sec- methodological chart applied in the SLR.
tor, 2) explore a variety of AI technology initiatives to achieve
sustainability growth objectives, and 3) analyse how agriculture firms/ 3. AI methods use in agriculture
farms improve sustainable growth through technologies which are
already underway and new technologies are being developed. This AI is one of the emerging areas of research in recent generation.
systematic review would contribute to enhance the understanding of Today AI is used to solve the problems particularly to reduce the use
the present view of the agriculture sustainability and agriculture of the labor force, to enhance efficient utilization of resources and to
technology. facilitate the development of sustainable business. With the rapid tech-
nological advancement,people are more intend to developed these ap-
2. Research methodology plications (Bannerjee et al., 2018). With that, different AI approaches
have been suggested to solve the existing problems in the agriculture
Systematic Literature Review (SLR) permits identifying and to improve the productivity.
obtaining relevant information on interesting subject area from the In our analysis we found that the main AI approaches used in
existing literature (Kitchenham and Charters, 2007). The SLR pursues agriculture are Neural Network (NN), DL, Fuzzy Logic, Support Vector

47
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Table 1
Agriculture functions and sustainability growth objectives.

Reference AI technology Technology used in agriculture function Agriculture function Sustainability

category objective

(Han et al., 2018) NN Prediction for agricultural output value Prediction 1

(Almomani, 2020) ANN Prediction model for agriculture waste Prediction 1
(Espejo-Garcia et al., 2020) DNN Weeds identification Weed control 1
(Yamaç, 2021) KNN, SVM, RF, AB Estimate sugar beet Etc for efficient irrigation management Prediction 1
(Mohapatra and Lenka, NN Crop monitoring Advanced care of crops 3
2016)
(Buyrukoğlu et al., 2021) ANN Prediction of Generic Escherichia coli population based on Prediction 1
Weather Station Measurements
(Dargan et al., 2020) Machine learning Sustainable agriculture supply chain performance Supply chain 2
applications
(Nguyen et al., 2019) ANN Agricultural landscapes management Resource management 4
(Liu et al., 2020) ANN Develop integrated agricultural drought index Prediction 1
(Castro et al., 2017) ANN High-performance prediction of Macauba fruit biomass Prediction 1
(Jung et al., 2021) AI Improve the resilience of agricultural systems Crop simulation Prediction 1
models utilize input variables such as crop management
information, weather, and soil data to estimate crop productivity
(Camaréna, 2020) AI Food production system Supply chain 2
(Zhang et al., 2021) LSTM Weather radar echo prediction method Prediction 1
(Dey and Shekhawat, 2021) AI Blockchain for sustainable e-agriculture Data management Supply chain 2
(Albalasmeh et al., 2020) ANN Predict the quality of the biochar based on operational conditions Prediction 1
of biochar production (parent biomass type, particle size,
pyrolysis temperature)
(Khan et al., 2020) Deep Learning Fruit Prediction Prediction 1
(Senocak and Goren, 2021) AI Forecasting the biomass-based energy potential Prediction 1
(Emmi et al., 2014) Robotics Integration and assessment of a real fleet Advanced care of crops 3
(Abdullahi et al., 2017) CNN Plant image recognition and classification Weed control 3
(McGuire, 2017) ANN Crop yield prediction Climate change impact assessment Prediction 1
(Guillén et al., 2021) Deep Learning Performance evaluation of edge-computing platforms for the Prediction 1
prediction of low temperatures
(Sharma et al., 2020) Machine Learning Applications for precision agriculture Advanced care of crops 3
(Espejo-Garcia et al., 2020) DNN Improving weeds identification Weed control 3
(Mohapatra and Lenka, ANN, Fuzzy Logic Pattern classification and weather dependent Fuzzy Logic Model Resource management 4
2016) for irrigation control
(Buyrukoğlu et al., 2021) ANN Prediction of Generic Escherichia coli Population in Agricultural Prediction 1
Ponds Based on Weather Station Measurements
(Giannakis et al., 2019) Cloud Environment Data sharing on production, diseases and weather Advanced care of crops 3
(Ellafi et al., 2021) ANN Prediction of saturated hydraulic conductivity (Ksat) in order to Prediction 1
enhance the efficacy of drainage system design in data-poor areas
utilizing existing and currently under-utilised datasets.
(Monteiro et al., 2021) ANN Weed control Weed control 3
(Santin et al., 2016) ANN Design of performance-oriented riparian buffer strips for the Advanced care of crops 3
filtering of nitrogen in agricultural catchments
(Taghavifar et al., 2015) ANN and Genetic Prediction of the power provided by the agricultural tractors Prediction 1
Algorithm
(Singh et al., 2012) ANN WPredicting sediment yield in the Nagwa agricultural watershed prediction 1
in Jharkhand, India
(Liu et al., 2021) ANN Predict rice growth rate Prediction 1
(Grimstad and From, 2017) Robotic Using cameras sensitive to visual and near infrared parts of the Advanced care of crops 3
electromagnetic spectrum to study plants
(Roshanianfard et al., 2021) Robotic arms and Seeding, watering, fertilizing, weeding, and harvesting Advanced care of crops 3
manipulation
systems
(Ireri et al., 2019) Machine learning Applied in real-time tomato post-harvesting procedures Low-cost Harvesting 1
tomato grading system based
(Birrell et al., 2020) Robotics Achieve a consistent harvesting cutting height, high-quality cuts Harvesting 1
(Mehta and Burks, 2016) Robotic Harvesting fruit detection efficiency, picking efficiency and Harvesting 1
picking rate
(Navas et al., 2020) Robotic Suitability of the cutting tools for the plants to be harvested Harvesting 1
(Navas et al., 2020) Robotic Suitability of the cutting tools for the plants to be harvested Harvesting 1
(Booth et al., 2020) Machine learning 3D estimate of the plant bulb's growth direction from a triplet of Advanced care of crops 3
2D x-ray images
(Raja et al., 2020) Robotic Crop signaling system Weed and crop classification Advanced care of crops, Weed 3
control
(Kounalakis et al., 2019) Deep learning Weed visual recognition algorithms Weed control 1
(Magalhães et al., 2021) Deep learning, Accurately identifying and detecting the mature fruit or fruit Harvesting 1
Harvesting robot bunches
(Khort et al., 2021) Robotic 10 h of continuous operation in low-light conditions in various Advanced care of crops 3
weather conditions.
(Aguiar et al., 2021) Deep Learning Detect tree trunks is still an area quite underdeveloped Advanced care of crops 3
(Vincent et al., 2019) Neural networks Agriculture land suitability analysis: Measurements of soil Resource management 4
and Multi-Layer moisture content, granular fragments (percentage of sand
Perceptron (MLP) particles in the soil), structure of the soil, compact and
cementation, cnternal drainage, available water content,
porousness, organic matter, cation exchange capacity, degree

48
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Table 1 (continued)

Reference AI technology Technology used in agriculture function Agriculture function Sustainability

category objective

saturation, pH value, salinity, and carbonates

(Hespeler et al., 2021) Robotic Harvesting in the evening hours or low light situations Harvesting 1
(Buzzy et al., 2020) Robotic Real-time leaf detection and counting Advanced care of crops 3
(Schor and Attwood-Charles, Robotic Disease detection and monitoring Advanced care of crops 3
2017)
(Navas et al., 2020) Robotic Weed treatment with a flaming and row crop cultivator Weed control 3
implement. Weed treatment with a herbicide patch sprayer. Pest
control with a canopy sprayer.
(Zapotezny-Anderson and Robotic Harvesting Harvesting 1
Lehnert, 2019)
(Kwon et al., 2019) Deep convolutional Fruit monitoring and grading systems Advanced care of crops 3
neural networks
(DCNNs)
(Zujevs et al., 2015) Robotic Fruit detection, localizing, gripping and picking Quality Advanced care of crops 3
measurements before picking
(Mendes et al., 2019) Robotic path AgRob Vineyard Detector
planning
Advanced care of 3
crops
(Paliwal et al., 2019) Robotic Soil data collection, disease detection, and field classification to Resource management 4
provide the best solutions for mixed cropping.
(Fue et al., 2020) Robotic Cotton harvesting Harvesting 1
(Linaza et al., 2021) Machine learning Yield prediction Prediction 1
(Bi et al., 2021) Slam robot Positioning system for agricultural environment Advanced care of crops 3
(Väljaots et al., 2018) Robotic Soil sampling and storage apparatus Resource management 3
(Jez et al., 2021) ANN, SVM, CNN Plant growth status, pest management, water and fertilizer Advanced care of crops 3
management for plant breeders and plant physiologists
(Porsch et al., 2019) Robotic Gantry pneumatic robotic manipulator for greenhouse Advanced care of crops 3
automation
(Zhang et al., 2021) Robotic Gripper developments to minimize the risk of damage to fruits, Harvesting 1
vegetables or food
(Jung et al., 2021) AI, Deep learning Irrigation management service Soil moisture monitoring system Resource management 4
to control irrigation, fight mildew, and deal with drought Image
recognition application to identify potential defects and nutrient
deficiencies in soil
(Kakani et al., 2020) Machine Learning Utilize the data collected from farms, irrigation, soil characteristics Resource management 4
and meteorological data to formulate field level insights as
recommendations for farmers to improve their overall yield
(Sharma and Bisen, 2013) Electric National e-NAM envisages spatial market integration, reduction in Prediction 2
Agriculture Market transaction costs and has direct implications on price signals and
(e-NAM) Deep price discovery, farmer's income and market liberalization
learning
(Oliveira et al., 2021) Robotic Robotic applications for land preparation, sowing and planting, Resource management Advanced 2
plant treatment, harvesting, yield estimation and phenotyping care of crops, Harvesting
(Beloev et al., 2021) Robotic Map or inspect a specific farming area in accordance to the Advanced care of crops 3
situation and the surrounding environment
(Isachsen et al., 2021) Robust robot-based Real-time speed and high registration accuracy and resolution Prediction 1
automation in enable the correct manipulation of food products without quality
primary production degradation
and processing
(Song et al., 2021) Robotic Greenhouse control system Advanced care of crops 3
(Thomopoulos et al., 2021) Robotic Kiwifruit harvesting robot Harvesting 1
(Seo and Umeda, 2021) Unmanned aerial UAVs are comparable to boom sprayers, showing similar Weed control 3
vehicles (UAVs) pest-control costs and management efficiency
(Ishii et al., 2021) Robotic Store, transport and relocate the boxes of tomato to the assigned Supply chain 2
storage area
(Peteinatos et al., 2020) CNN Plant and weed classifications Weed control 3
(Fahey et al., 2021) AI-based data fusion Identify and quantify disease and pest epidemics accurately and at Advanced care of crops 3
technique the earliest possible stage
(ÖZLÜOYMAK et al. Robotic Weed control system Weed control 3
Özlüoymak et al., 2019)
(Balafoutis et al., 2017) Machine learning Autonomous plant classification Advanced care of crops 3
(Spanaki et al., 2021) AI AgriTech drones Collecting data from the fields, and support monitored human Prediction 1
decision making for everyday tasks (e.g. disease inspection, crop
monitoring) and AgriFood operations (e.g. irrigation, fertilization
etc.) of the farm
(Bi et al., 2020) Deep Learning Predict consumer Yogurt preferences based on sensory attributes Prediction 2
(Kiourt et al., 2020) Deep Learning Automatic image-based food recognition Advanced care of crops 3
(Dargan et al., 2020) Deep Learning Predict wine taste preference Prediction 2
(Chukkapalli et al., 2020) AI Smart farming Cooperative ecosystem Advanced care of crops 3
(Utstumo et al., 2018) Robotic Drop-on-Demand (DoD) weed control system Weed control 3
(Lytridis et al., 2021) Robotic Land preparation Resource management 4
(Hossain and Komatsuzaki, Robotic Weed management Weed control 3
2021)

(continued on next page)

49
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Table 1 (continued)

Reference AI technology Technology used in agriculture function Agriculture function Sustainability

category objective

(Kultongkham et al., 2021) Robotic Tomato harvesting Harvesting 1

(Ruigrok et al., 2020) Robotic Weed detection Weed control 3
(Magomadov, 2019) Deep Learning Plant disease detection Advanced care of crops 3
(Feng et al., 2018) Robotic robotic harvesting system for cherry tomato Harvesting 1
(Grieve et al., 2019) Robotic Weed control Weed control 3
(Midtiby et al., 2016) Robotic weeding application Weed control 3
(Williams et al., 2019) Machine Vision, Kiwifruit Harvesting Harvesting 1
Convolutional
Neural Networks,
and Robotic
(Hespeler et al., 2021) Deep learning harvesting of chili peppers Harvesting 1
(Gonzalez-de Santos et al., Robotic weed and pest control Weed control 3
2017)
(Sudars et al., 2020) Robotic computer Annotated food crops and weed images Weed control 3
(Ngugi et al., 2021) Machine learning leaf pest and disease recognition Advanced care of crops 3
(Ghafar et al., 2021) Robotic Spraying fertilizers and pesticides Weed control 3
(Azmi et al., 2021) Robotic Crop seeding Harvesting 1
(Yorozu et al., 2021) Robotic Smooth and safe harvesting support in the field Advanced care of crops 3
(Kim et al., 2021) Robotic Estimate crop height and detect the target crop region Prediction 1
(Gai et al., 2021) Robotic Generating crop field maps as occupancy grids and providing Prediction 1
inter-row vehicle positioning data
(Panarin and Khvorova, Robotic Taking into account the physical environment conditions and Prediction 1
2021) build mathematical models
(Rysz and Mehta, 2021) Robotic Fruit harvesting Harvesting 1
(Zangina et al., 2021) Robotic Selective and variable spray of pesticides to the plants Weed control 3
(Mohamed et al., 2020) Machine learning Spatial mapping analysis of soil characteristics Resource management 4
(Gupta et al., 2020) Deep learning Soil parameters analysis Resource management 4
(Mohammed and Jassim, Robotic Seeding, fertilization and initial irrigation process Harvesting, Advanced care of crops 1,3
2021)
(Villa-Henriksen et al., Robotic Harvesting Harvesting 1
2021)
(Ünal et al., 2021) Robotic Soil penetration resistance and electrical conductivity Resource management 4
(Li et al. 2021a) Deep learning Weed detection Weed control 3

Machine (SVM), Random Forest, K-nearest and Robotics. As in Fig. 2 cameras. This process assists to detect the presence of pests in the
from the selected papers the most of researchers are applied robotics crop field. This problem is solved using low-cost agricultural robot
which is 44%. Thereafter, NN and DL is 26% and 15%, respectively. (Azmi et al., 2021). For agricultural cyber-physical systems, researchers
Robotics are used in agriculture to assist farmers. These robots are proposes a suggest an intelligent management deign using robotic
developed with many operations such as weeding application, visual technique(Huang et al., 2021). The robots can use in labour intensive,
detection and harvesting where they can be used to match the needs repetitive and physical demanding tasks in agricultural field. The recent
of various tasks (Zhang et al., 2020; Benos et al., 2020; Yorozu et al., literature reveals that robots are being used to perform several special-
2021). Ghafar et al. (Ghafar et al., 2021) design a robot to spray pesti- ized tasks which were performed by experience farmers (Marinoudi
cides and fertilizers in harvesting field at low operating cost and general et al., 2019; Le et al., 2019; Zhang and Noguchi, 2017; Huang and
crop monitoring. This model is used two-wheeled robot that included a Chang, 2019; Kim et al., 2021). As such, there are advantages using ro-
a mobile base which is used a spewing mechanism with a controlling of botics in agriculture such as production increase, widening the profit
wireless tool that is used to manage the movements of the robot. There- and saving time for performing repetitive tasks. It is estimated that pes-
after, crop growth conditions and health factors are monitored using ticides usage can be reduced by 80% if the farmers use robots to spray

Fig. 1. Methodological chart.

50
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Fig. 2. AI approaches used in agriculture.

fertilizers. Moreover, since robots cab work around trees, rocks, lakes nitrogen content in soil and leaf, detect irrigation and plants' water
and other obstacle areas easily, crops can be cultivated more fields stress level, assess erosion of water, detect usage of herbicide, defects
(Khare et al., 2021; Gai et al., 2021; Panarin and Khvorova, 2021; Rysz on food and damage of crop hail and monitor greenhouse (Bu and
and Mehta, 2021). Wang, 2019; Li et al. 2021c; Zhou et al., 2021; Chen et al., 2020; Xue
Artificial neural networks (ANN) are one of the most important et al., 2019). However, DL models need comprehensive datasets as the
technique of AI. These models are developed using interconnected input to serve at the training procedure. Other than the above methods,
nodes which are performed functions as our human brain. The usage researcher were also applied Fuzzy Logic, SVM, Random Forest and
of NNs application is very wide, and it also includes in agriculture K-nearest (Kurniasih et al., 2018; Center and Verma, 1998; Pujari
(Kujawa and Niedbala, 2021). (Almomani, 2020) was applied a NN to et al., 2016; Jez et al., 2021; Yamaç, 2021). However, Robotic models
optimize the cumulative methane production. This NN model has and DL models have showed significant usage in the agriculture field.
showed significant results in prediction. In addition, NN has employed To develop a more complete understanding of the enabling AI tech-
to pattern classification and soil moisture content prediction nologies currently applied in agriculture sector, Fig. 3 further illustrates
(Mohapatra and Lenka, 2016). Scaled Conjugate Gradient and BFGS main AI approaches used in agriculture functions.
Quasi-Newton based neural network algorithms used to take various Moreover, the study aligns the main agriculture functions with
soil and environmental parameters and predict hourly requirement of sustainability growth objective (Fig. 4) in order to provide contextual
soil soil moisture content. The NN has become popular as a classification link with present view of the agricultural functions and sustainability.
method in agricultural engineering. NNs are virtuous to formulate the In here, the study focused on several functional areas highlighted in
model using non-linear data and data represents with images. the selected studies. As such, the research hots-pots of AI and
Therefore, this approach is good for crop classification using image agriculture in the past decades comprise mainly prediction, harvesting,
data (Boniecki et al., 2020). The prediction of growth rate of rice is im- advanced care crops, weed control, resource management and supply
portant to obtain sustainability in agriculture. Researchers recom- chain.
mended rice growth rate modeling using NN which shows less errors
compared to regression algorithm and gene expression programming 4. Agriculture functions and AI
(Liu et al., 2021).
With the training limitation of NN, researchers are used DL. DL deals This section elaborates the inclusive review of the literature that
with recent and modern technique to process images and analyse data, applied automate functions in agriculture (e.g., prediction, harvesting,
which guarantees the potential results. Application of DL into agricul- advanced care crops, weed control, resource management and supply
tural domain is emerged instead of various domain that DL has been chain) using AI techniques. Within the reviewed papers it was identified
successfully applied (Kamilaris and Prenafeta-Boldú, 2018; Zhu et al., that the most common applications of AI are predicting, harvesting,
2018; Santos et al., 2019; Nguyen et al., 2020). Jiang et al. (Jiang et al., advanced care of crop and so on Fig. 5.
2021) suggest a method to identify the disease in fruit like Apple and
the method is useful to prevent the disease without harming the 4.1. Prediction
environment. In the method, capability of image processing and classi-
fication in DL were applied to classify the fruit image. Deep neural As shown in Fig. 5, of the total 115 studies, 36 (40%) articles reported
network with different convolution layers and different number of neu- that the most common applications of AI for agriculture is prediction
rons are examined and evaluated. The results beat the performance of model for total agricultural output value (Tian et al., 2021; Kim et al.,
baseline models. DL stimulates multi-model approach to detect, dissem- 2021; Han et al., 2018; Khan et al., 2020; Crane-Droesch, 2018; Kumar
inate and monitor the Active Fire Locations (AFL) in agricultural tasks and Joshi, 2015; Isachsen et al., 2021), waste minimization (Almomani,
and they are guaranteed the highly accurate results (Sharma et al., 2020), irrigation control (Mohapatra and Lenka, 2016), weather index
2021). We found that DL has applied to identify seeds and pest, monitor (Buyrukoğlu et al., 2021; Liu et al., 2020; Zhang et al., 2021), energy

51
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Fig. 3. Main AI approaches used in agriculture functions.

optimization (Senocak and Goren, 2021; Guillén et al., 2021) and de- biomass, and land area to estimate yield, manage irrigation and land
mand and consumer preference (Sharma and Bisen, 2013; Giannakis area and develop drought index. The fact that this AI application in pre-
et al., 2019; Dargan et al., 2020; Monteiro et al., 2021). To cover these sce- diction model in agriculture is so common can be justified by the com-
narios, ANN and DL techniques are used by 95% of the reviewed papers. plex and dynamic nature of the agricultural parameters, thus, it is
The ANN and DP are used to collect real-time data about multiple agricul- perplexing to obtain accurate predictions. Thus, AI can be served as a
tural parameters, such as production quantity, waste, climate data, method to face the complexities in the dynamic nature of agricultural

Fig. 4. Agriculture functions and Sustainability growth objectives.

52
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Fig. 5. AI usage in agriculture functions.

parameters. Interestedly, few papers (4) focused on predicting customer certain advantages such as minimum wastage, picking efficiency, high
demand and preference and market integration for few agricultural picking rate and flexible work force and nighttime operation. Develop-
products (e.g.: wine, yogurt). The studies focused on the the elements ment of timely, efficient, and careful robotic harvesting solutions lead
in sensory functions and reveal the associations among evaluators scor- to complete the harvesting process while generating high quality yields
ing and latent features. These preference predictions provide wider at minimum time consuming and at minimum unrecuperative damages
chances to develop food product designs in future. What we can con- in the harvesting process. Particularly, there have been significant
clude from these literature base is that agriculture sustainability is en- developments of AI towards the sustainability agriculture objective of
sure production of an adequate food supply. ensuring production of an adequate food supply. Besides research
projects have been performed, very few have developed into the com-
4.2. Harvesting mercial world (Kiwi fruit; Tomato; Cotton; Apple; Rice).

Harvesting is a challenging task in agriculture because harvesters 4.3. Advanced care of crop
strongly correlate to crop detection, quality cuts, damage, picking and
packaging. The labour assets utilize for harvesting is one of the main It is essential to repetitive detection and monitoring on the plant's
cost components in agriculture production. To overcome the high la- life cycle in order to attain the yield with high quality and quantity.
bour cost component, the prior studies pay wider attention to exploit The plant growth and development could be detected with the number
commercially doable AI applications for harvesting. As presented in of leaves and that would be the key phenotype of plant growth and crop
Fig. 5, the next most common application of AI is harvesting (28 articles; damage by attacks of bacteria, fungi, and other pests are threatening the
31%). Moreover, robotic technique (robot arm) has been identified as an long-term viability of plant phenotyping. Thus, advanced care of crops is
effective tool for assistance in the agriculture industry to harvest the another agricultural function where AI is mostly applied (26 articles;
fruit/vegetable while not harming the plant (Hespeler et al., 2021). As 29%). AI technologies, such as DL, ANN, robotic and ML, provide the
a functional model, the robot could automatically move on the rail, means to automate disease detection, measure plants, monitoring
identify, detect and locate the mature bunch, hold and separate the tar- plant growth status and applying fertilizer (Zargar et al., 2020; Emmi
get and collect the crop harvested. Precisely identifying and detecting et al., 2014; Sharma et al., 2020; Santin et al., 2016; Grimstad and
the mature crops encompass a key technique of harvesting robot. An ef- From, 2017; Raja et al., 2020; Magalhães et al., 2021; Buzzy et al.,
ficient object detection and inspection algorithm are necessary for a 2020; Schor and Attwood-Charles, 2017; Santos et al., 2020; Li et al.
robotic platform to be used in harvesting. Different types of sensor tech- 2021b; Yorozu et al., 2021). Symptoms of diseases developed by attacks
nologies use to detect and locate crop in the tree branches, determine of bacteria, fungi, and other pests need to be identified in an initial stage
the ripeness of the crop, determine the geometry of the tree canopy according to the changes in the physiological condition of plant parts
and locate the tree in orchard, and finally to pick the crop from the (leaves, stems, and flowers) to provide treatment at the right time.
tree. Computer vision is used for crop ripeness estimation (Hespeler Currently, labour intensive crop caring practices use vast amount of
et al., 2021). In all cases, crops should be picked when they are ripe or agricultural chemical inputs (fertilizers, herbicides, fungicides, and in-
mature without mechanical damage to the fruit. This action should secticides) cause to have high production cost and lead pollution mat-
take place as be as quick and as cost-effective as possible. Thus, research ters as well. In general, it is estimated that more than 100 kg are
on harvesting robot mainly concentrations on five key areas; identifying applied per hectare in farm land. Unfortunately, majority of the nitrate
targets under complex background; separating soft crop; level of con- applied were either washout or loss in the air. The robotic disease-
suming energy to harvest, harvesting tools suitability and conformation detection systems were commonly designed in whole inclusive pattern
design to fit with unshaped work fields (Navas et al., 2020). Moreover, to identify the results in infection and these results could be utilised
thermal imaging for real-time harvesting robot allow to harvest in the to detect precise diseases and and apply fertilizes appropriately
evening hours or low light situations (Hespeler et al., 2021). Finally, (Schor and Attwood-Charles, 2017; Grimstad and From, 2017). In
vision-based crop detection is a critical component for robotic harvest- addition, ML technique was used to measure plants with sensors
ing and it includes crop detection with dimension, mass estimation, and (Sharma et al., 2020) to estimate plant growth direction (Booth et al.,
localization prior to pick or slice (Lee et al., 2020; Zujevs et al., 2015; 2020) and plant classification (Libertn et al., 2018), ANN to design
Villa-Henriksen et al., 2021). Using a robotic system would enable performance-oriented riparian buffer strips for the filtering of nitrogen

53
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

(Santin et al., 2016) and measure plant growth status (Li et al. 2021b), reach final product the end consumers. In the process of SC also includes
DL to detect underdeveloped plants (Aguiar et al., 2021) and to recog- multiple stakeholders such as farmers, producers, processors, certifica-
nize the plant using image-based (Kiourt et al., 2020). Therefore, it is tion agencies, traders, government, retailers, distributors, and final
important to recognize the AI applications that favor both disease man- consumers. Compared with other supply chains, agriculture SC is
agement and to provide sufficient, safe, and nutritious food to the global complex due to the nature of perishability and high supply-demand
population. Particularly, there have been significant developments of AI fluctuations of the products and high consumer awareness towards pro-
towards the sustainability agriculture objective of achieving better duce provenance, quality, and safety. All these notes, the review insists
health and nutrition for a growing population. that AI applications, especially ML, in agriculture SC enable farmers and
other relevant organizations to draw valuable insights on agriculture
4.4. Weed control process, leading to increase agricultural productivity while taking
decisions via data-driven platform (Sharma et al., 2020; Camaréna,
It is necessary to control weed in the crop field to increase the produc- 2020). Data pays a crucial role in supply chains thus improvisation in
tion of agriculture. The review (26 articles; 29%) highlights that AANN, storage, collection, visualization, privacy, security, accuracy, and access
DNN, CNN and DL to identify and classify the plants as weed using of agriculture data can impact application of AI in agriculture supply
image processing. These techniques include the crop signaling compound chain. The common believe in agriculture is that farmer is classified
includes distinctive characteristics that assure the detection of crop dis- under low income group and many firms/farms in the agricultural are
eases and ensure the classification of crop and weed (Zangina et al., worried with low-profitability. AI is widely used in SC to identify hidden
2021; Espejo-Garcia et al., 2020; Abdullahi et al., 2017; Monteiro et al., patterns in the data, in this line, SC stakeholders consider to accelerating
2021; Raja et al., 2020; Kounalakis et al., 2019; Peteinatos et al., 2020). AI in SC, leads to achieve the expectations of farmers as well as cus-
Next, the signal is transmitted to the robotic arm or Unmanned aerial ve- tomers. Contextual factors that have been identified as important
hicles (UAVs) to pluck the plant through serial communication or execute influencers of AI in agriculture sustainability objective: alleviate poverty
weed treatment with a flaming and row crop cultivator implement, weed of farmers through formalize products sales to certified markets and
treatment with a herbicide patch sprayer or canopy sprayer (Emmi et al., global commodity price trends, visualize the farm income prior to the
2014; Seo and Umeda, 2021; Özlüoymak and Bolat, 2020; Kounalakis intervention and the formalize the existence market structure. As re-
et al., 2019). Using AI in weed control enables to decrease unnecessary view highlights although SC using AI platforms leads to sustainable ag-
plants within fewer time frames and minimize fertilizers and herbicides riculture objective, questions related to the mechanism of reaction and
utilization, which cause soil degradation and pollution. selectivity of matrices for AIs in consumer aspect are still unanswered.

4.5. Resource management 5. Considerations

Agriculture is naturally bounded with the resource constraints Fig. 6 tabulates the contextual link of usage of AI techniques that are
(e.g., land, water and soil). Primarily,womb of agriculture is soil and supposed to address sustainable agriculture.
soil management therefore serves as primal concern in agricultural re- The number of studies that address sustainable agriculture is
source management. Thus, assessing the suitability of agricultural land increasing; however less attention has been devoted to investigating
becomes the vital task in agriculture development. Moreover, in preci- the sustainability aspect of agriculture with regard to AI technology. Ob-
sion agriculture, irrigation management plays a crucial role. The review viously, the sustainable agricultural is not a typical research field for AI
emphasized that AI driven agriculture is focusing on methods to opti- researchers to study; the field of AI research has a historical practice
mize land (Nguyen et al., 2019; Oliveira et al., 2021), soil (Nguyen to focus on the industries that involving with new products and ser-
et al., 2019; Moya-Rico et al., 2019; Paliwal et al., 2019; Väljaots et al., vices. Furthermore, researches related to agricultural food tend to ex-
2018) and water/irrigation (Mohapatra and Lenka, 2016; Jung et al., amine the way of increasing production rather that addressing
2021; Kakani et al., 2020) considering the benefit it brings to people sustainability issues. Moreover, most research within the agriculture
linked with this profession. Moreover, the weather forecasts such as food industry tends to examine production rather than sustainable is-
sunlight, rainfall, humidity, and moisture guide by using AI leads to sues. Due to the vast and increasingly expanding body of literature on
the optimal use of water for scheduling and planning the crop. To agri-technology, this review has focused on how AI technology can
cover these scenarios, ANN, DL, ML and robotic techniques are widely improve the sustainability of agriculture industry. The aim of this review
used by the reviewed papers. Ground robots and UAVs are more paper is therefore to analyse and create an understanding of the
precisely used to collect soil and water sample and land preparation/ different types of AI applications in agriculture industry and how
sowing. Neural networks, deep and ML techniques used to computed those applications align to achieve the agriculture sustainability objec-
the normalized soil moisture index to estimate the soil moisture content tives. A systematic and quantitative evaluation of different agricultural
and develop a model to assess the agriculture land for cultivation in parameters is of vital importance to improve agriculture production
terms of four decision classes, namely more suitable, suitable, and ensure sustainable food supply. Unsustainable agricultural produc-
moderately suitable, and unsuitable. Since irrigation management tion practices such as food wastage and production shocks due to cli-
plays a critical role in quantity and quality of the crops, estimating mate changes can be minimized if the sector uses AI to get accurate
evapotranspiration, streamflows and real-time management of reser- predictions. Our review therefore underscores the importance of AI pre-
voir release by using ML algorithms are highlighted in the review diction driving the adoption of modern agricultural innovations to en-
(Sharma et al., 2020). ML helps to process all data samples to construct sure adequate food production and supply. Within the context of the
a heuristic model that can predict factors resulting high yields. While different concepts of sustainable agriculture objectives, AI applications
the use of UAVs and robots for sowing has advantages like large area in prediction specifically focus on the potential contribution towards
coverage and speed. However, uncertainty in ground measurements satisfying human needs for food. However, the vast majority of agricul-
and power requirements are restricting the number of task they can ture products remain unaddressed, and almost no fully automated
perform (Lytridis et al., 2021). prediction models have been developed. Moreover, developing a
specific prediction model for market demand and preference of agricul-
4.6. Supply chain ture products is insufficient. However, due to the low repeatability and
difficulties in corresponding, AI implementation in agriculture sector
Supply chain (SC) in agriculture includes several tasks such as pre- become main challenge (Linaza et al., 2021) specifically in developing
production, production, storage, processing, distribution, retail, and nations, which requires immediate solution/s.

54
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Fig. 6. AI techniques that are supposed to address sustainable agriculture.

Development of timely, efficient, and careful robotic harvesting natural resources using AI technologies lead to increased global crop
solutions lead to obtain desired quality harvest at short-time period yields. Meeting the fourth objective - conserving natural resources, is a
and more interestedly at minimum unrecoverable loss. Particularly, daunting challenge. As we identified, it includes protection of soil health
there have been significant developments of AI towards the sustainabil- and water quality and maintain biodiversity of flora, fauna, and natural
ity agriculture objective of ensuring production of an adequate food landscapes. The review sheds a light that AI technological developments
supply. Besides research projects have been performed, very few have taking consideration on these challenges while improving agri-food
developed into the commercial world (Kiwi fruit; Tomato; Cotton; productivity with minimum effects to the environment. Particularly,
Apple; Rice). Moreover, the review revealed that disease detection sys- there have been significant developments of AI towards the sustainabil-
tems are mostly concerned on leaves and the lower area of leaves is not ity agriculture objective of conserving natural resources. However, to
properly sensed by the camera sensor. As said, besides leaves, AI tech- improve overall crop water productivity, AI technology has to be ad-
niques need to be expanded for a variety of plant parts. In line with that vanced in irrigation technologies such as efficient low pressure center
(Li et al. 2021b) insisted that AI technologies have been counted on a dif- pivot irrigation and micro-irrigation and weather-based and soil mois-
ferent sensors and imaging technologies to gather a variety of plant data, ture sensor-based irrigation scheduling. Moreover, improving decision
whereas analyzing different sensor data also relies on different hardware support tools integrating weather, soil and crop information will ensure
devices, software systems, and different platforms and different monitor- progress towards the sustainable objectives of agriculture.
ing scales used to analyse date. Such complex operation process might Though AI helps to enhance the visibility of agriculture SC, more at-
slow down data acquisition and integration, leading to an information tention need to be focused on the food retailing phase for predicting
lag. This implies the need for international harmonization and standardi- consumer demand, perception and buying patterns. A precise prediction
zation in phenotyping data. The phenotyping data collecting and analyz- of food requirements or food consumption behavior of buyers helps to
ing then could lead to manage cultivating practices, plant breeding and avoid overstocking, overproduction, resources overutilization and guaran-
overall management in agricultural functions. tee the fair income and price to farmers and buyers respectively. Dey et al.
Weeds destructively affect agricultural crop productions by (Dey and Shekhawat, 2021) enhance the back and forward linkages in
contending with crop plants for resources, including soil moisture and supply chain, reduce transportation cost and delivery time, enhance
nutrients. Providing sufficient and healthy foods for ever growing farmers' awareness on price, selling quotas, available stocks and online
population heavily depends on the ways we control weeds and apply showcases and reduce the risks involved in contract arrangement. Subse-
fertilizers in efficient manner. For future work, the review opines quently, investments in AI applications in agriculture industry have exhib-
(Kounalakis et al., 2019) that more robust weed recognition approaches ited the possibility of achieving four objectives in agriculture sustainability
could be extended with additional data capturing conditions (like illu- while enhancing the farmers' livelihood, minimizing food production cost,
mination, grass density) and sampling techniques could be synthetized controlling food price fluctuation and ensuring food choices to consumers.
with techniques like Synthetic Minority Over-sampling Technique It is observable that AI application in achieving the second sustainable
(SMOTE) and Adaptive Synthetic Sampling (ADASYN). objective – alleviate poverty of farming community, remains scantly
Natural resources were jeopardized and different forms of environ- addressed. On this note, there is a vital requirement to design comprehen-
mental degradation became apparent, thus conservation practices of sive framework of AI that should be used in agriculture SC.

55
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

6. Conclusion and further research directions Aguiar, A.S., Magalhães, S.A., Dos Santos, F.N., Castro, L., Pinho, T., Valente, J., Martins, R.,
Boaventura-Cunha, J., 2021. Grape bunch detection at different growth stages using
deep learning quantized models. Agronomy 11, 1890. https://doi.org/10.3390/agron-
As sum in the literature consulted, we observed that AI applications omy11091890.
are extensively adopted nowadays to enhance operational automation Ala-Harja, H., Helo, P., 2015. Reprint of “green supply chain decisions-case-based perfor-
mance analysis from the food industry”. Transp. Res. Part E: Logist. Transp. Rev. 74,
and performance of agricultural industry. We found that the most 11–21. https://doi.org/10.1016/j.tre.2014.12.005.
common applications of AI for agriculture are prediction model for Albalasmeh, A., Gharaibeh, M.A., Mohawesh, O., Alajlouni, M., Quzaih, M., Masad, M., El
total agricultural output value, followed by harvesting applications. Hanandeh, A., 2020. Characterization and artificial neural networks modelling of
Though agriculture is naturally bounded with the resource constraints, methylene blue adsorption of biochar derived from agricultural residues: effect of
biomass type, pyrolysis temperature, particle size. J. Saudi Chem. Soc. 24, 811–823.
AI applications in natural resources management (such as water, soil, https://doi.org/10.1016/j.jscs.2020.07.005.
land) are presently at unsatisfactory level. We were further able to iden- Alexandratos, N., Bruinsma, J., 2012. World agriculture towards 2030/2050: the 2012
tify that final consumer aspect in agriculture SC needs to be devouring revision. ESA Working Papers 12-03 https://doi.org/10.22004/ag.econ.288998.
Almomani, F., 2020. Prediction of biogas production from chemically treated co-digested
extensive attention in order to achieve one of the key agriculture sus- agricultural waste using artificial neural network. Fuel 280, 118573. https://doi.org/
tainability objective; alleviate poverty of farmers. Moreover, it was 10.1016/j.fuel.2020.118573.
witnessed in this review that in recent work the use of AI and image Åstrand, B., Baerveldt, A.J., 2002. An agricultural mobile robot with vision-based percep-
processing techniques has become more common to improve the sus- tion for mechanical weed control. Auton. Robot. 13, 21–35. https://doi.org/10.1023/
A:1015674004201.
tainable agriculture. When consider the DL models, these models are Azmi, H.N., Hajjaj, S.S.H., Gsangaya, K.R., Sultan, M.T.H., Mail, M.F., Hua, L.S., 2021. Design
suffer from task dependent since all the models are using general and fabrication of an agricultural robot for crop seeding. Mater. Today: Proc. https://
word embedding vector. To overcome this problem attention-based doi.org/10.1016/j.matpr.2021.03.191.
Balafoutis, A.T., Beck, B., Fountas, S., Tsiropoulos, Z., Vangeyte, J., van der Wal, T., Soto-
DL models can be developed. In the future researchers can consider de-
Embodas, I., Gómez-Barbero, M., Pedersen, S.M., 2017. Smart farming technologies–
veloping attention-based DL models. Likewise, the future studies should description, taxonomy and economic impact. Precision Agriculture: Technology and
intend to address the gaps that identified in this systematic review such Economic Perspectives. Springer, pp. 21–77 https://doi.org/10.1007/978-3-319-
68715-5_2.
as under-utilised commercial crops appraisal, data capturing conditions,
Bannerjee, G., Sarkar, U., Das, S., Ghosh, I., 2018. Artificial intelligence in agriculture: a lit-
natural resource standards, functional areas and geographical locations. erature survey. Int. J. Sci. Res. Comp. Sci. Appl. Manag. Stud. 7, 1–6.
This study has provided useful information to understand the impli- Barth, H., Ulvenblad, P., Ulvenblad, P.O., Hoveskog, M., 2021. Unpacking sustainable busi-
cation of AI in agriculture sustainability. However, there can be limita- ness models in the swedish agricultural sector–the challenges of technological, social
and organisational innovation. J. Clean. Prod. 304, 127004. https://doi.org/10.1016/j.
tions in this research. This work may extend by considering project jclepro.2021.127004.
costs, usability and regional challenges in AI applications. In addition, Beloev, I., Kinaneva, D., Georgiev, G., Hristov, G., Zahariev, P., 2021. Artificial intelligence-
can explore how attention-based DL models are used in agriculture driven autonomous robot for precision agriculture. Acta Technol. Agric. 24, 48–54.
https://doi.org/10.2478/ata-2021-0008.
with the newest AI improvements. Due to the growing application of
Benos, L., Bechar, A., Bochtis, D., 2020. Safety and ergonomics in human-robot interactive
AI itself is not adequate to obtain sustainable objectives; it is required agricultural operations. Biosyst. Eng. 200, 55–72. https://doi.org/10.1016/j.
to assess the adaptability of AI together with other useful maneuvers biosystemseng.2020.09.009.
like policy support for AI developers and programme intervention to Bi, K., Qiu, T., Huang, Y., 2020. A deep learning method for yogurt preferences prediction
using sensory attributes. Processes 8, 518. https://doi.org/10.3390/pr8050518.
implementations.
Bi, X., Tang, W., Gao, Z., 2021. A positioning system on robot for agricultural environment.
Journal of Physics: Conference Series. IOP Publishing, p. 042024. https://doi.org/10.
Credit authorship contribution statement 1088/1742-6596/1865/4/042024.
Birrell, S., Hughes, J., Cai, J.Y., Iida, F., 2020. A field-tested robotic harvesting system for ice-
berg lettuce. J. Field Robot. 37, 225–245. https://doi.org/10.1002/rob.21888.
Vilani Sachithra: Conception and design of study, Acquisition of data, Bolis, I., Morioka, S.N., Sznelwar, L.I., 2014. When sustainable development risks losing its
Analysis and/or interpretation of data, Writing –original draft, Writing – meaning. Delimiting the concept with a comprehensive literature review and a con-
review & editing L.D.C.S. Subhashini: Conception and design of study, ceptual model. J. Clean. Prod. 83, 7–20. https://doi.org/10.1016/j.jclepro.2014.06.041.
Boniecki, P., Koszela, K., Świerczyński, K., Skwarcz, J., Zaborowicz, M., Przybyl, J., 2020.
Acquisition of data, Analysis and/or interpretation of data, Writing –
Neural visual detection of grain weevil (sitophilus granarius l.). Agriculture 10, 25.
original draft, Writing – review & editing. https://doi.org/10.3390/agriculture10010025.
Booth, B.G., Sijbers, J., De Beenhouwer, J., 2020. A machine learning approach to growth
Declaration of Competing Interest direction finding for automated planting of bulbous plants. Sci. Rep. 10, 1–11.
https://doi.org/10.1038/s41598-019-57405-8.
Bu, F., Wang, X., 2019. A smart agriculture iot system based on deep reinforcement learn-
The authors declare that they have no known competing financial ing. Futur. Gener. Comput. Syst. 99, 500–507. https://doi.org/10.1016/j.future.2019.
interests or personal relationships that could have appeared to influ- 04.041.
Buyrukoğlu, G., Buyrukoğlu, S., Topalcengiz, Z., 2021. Comparing regression models with
ence the work reported in this paper. count data to artificial neural network and ensemble models for prediction of generic
escherichia coli population in agricultural ponds based on weather station measure-
Appendix 1. AI, Agriculture functions and Sustainability growth ments. Microb. Risk Anal., 100171. https://doi.org/10.1016/j.mran.2021.100171.
objectives Buzzy, M., Thesma, V., Davoodi, M., Mohammadpour Velni, J., 2020. Real-time plant leaf
counting using deep object detection networks. Sensors 20, 6896. https://doi.org/
10.3390/s20236896.
Sustainable objectives Camaréna, S., 2020. Artificial intelligence in the design of transition to sustainable food
systems. J. Clean. Prod. 271, 122574. https://doi.org/10.1016/j.jclepro.2020.122574.
1. Ensure production of an adequate food supply. Castro, N.R., Swart, J., 2017. Building a roundtable for a sustainable hazelnut supply chain.
2. Alleviate poverty. J. Clean. Prod. 168, 1398–1412. https://doi.org/10.1016/j.jclepro.2017.08.239.
Castro, C.A.D.O., Resende, R.T., Kuki, K.N., Carneiro, V.Q., Marcatti, G.E., Cruz, C.D., Motoike,
3. Achieve better health and nutrition for a growing population.
S.Y., 2017. High-performance prediction of macauba fruit biomass for agricultural and
4. Conserve natural resources. industrial purposes using artificial neural networks. Ind. Crop. Prod. 108, 806–813.
https://doi.org/10.1016/j.indcrop.2017.07.031.
Center, B., Verma, B.P., 1998. Fuzzy logic for biological and agricultural systems. Artificial
References Intelligence for Biology and Agriculture. Springer, pp. 213–225. https://doi.org/10.
1007/978-94-011-5048-4_11.
Abdullahi, H.S., Sheriff, R., Mahieddine, F., 2017. Convolution neural network in precision Chen, H., Chen, A., Xu, L., Xie, H., Qiao, H., Lin, Q., Cai, K., 2020. A deep learning cnn archi-
agriculture for plant image recognition and classification. 2017 Seventh tecture applied in smart near-infrared analysis of water pollution for agricultural irri-
International Conference on Innovative Computing Technology (INTECH). IEEE , gation resources. Agric. Water Manag. 240, 106303. https://doi.org/10.1016/j.agwat.
pp. 100–125. https://www.researchgate.net/profile/Halimatu-Abdullahi/public 2020.106303.
ation/320971643_Convolution_neural_network_in_precision_agriculture_for_plant_ Chukkapalli, S.S.L., Mittal, S., Gupta, M., Abdelsalam, M., Joshi, A., Sandhu, R., Joshi, K.,
image_recognition_and_classification/links/5a1d4fee4585153731898423/Convol 2020. Ontologies and artificial intelligence systems for the cooperative smart farming
ution-neural-network-in-precision-agriculture-for-plant-image-recognition-and- ecosystem. IEEE Access 8, 164045–164064. https://doi.org/10.1109/ACCESS.2020.
classification.pdf. 3022763.

56
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Crane-Droesch, A., 2018. Machine learning methods for crop yield prediction and climate Huang, C.H., Chen, P.J., Lin, Y.J., Chen, B.W., Zheng, J.X., 2021. A robot-based intelligent
change impact assessment in agriculture. Environ. Res. Lett. 13, 114003. https://doi. management design for agricultural cyber-physical systems. Comput. Electron.
org/10.1088/1748-9326/aae159. Agric. 181, 105967. https://doi.org/10.1016/j.compag.2020.105967.
Dargan, S., Kumar, M., Ayyagari, M.R., Kumar, G., 2020. A survey of deep learning and its Ireri, D., Belal, E., Okinda, C., Makange, N., Ji, C., 2019. A computer vision system for defect
applications: a new paradigm to machine learning. Archiv. Comp. Methods Eng. 27, discrimination and grading in tomatoes using machine learning and image process-
1071–1092. https://doi.org/10.1007/s11831-019-09344-w. ing. Artific. Intellig. Agric. 2, 28–37. https://doi.org/10.1016/j.aiia.2019.06.001.
Dey, K., Shekhawat, U., 2021. Blockchain for sustainable e-agriculture: literature review, Isachsen, U.J., Theoharis, T., Misimi, E., 2021. Fast and accurate gpu-accelerated, high-
architecture for data management, and implications. J. Clean. Prod., 128254. https:// resolution 3d registration for the robotic 3D reconstruction of compliant food objects.
doi.org/10.1016/j.jclepro.2021.128254. Comput. Electron. Agric. 180, 105929. https://doi.org/10.1016/j.compag.2020.
Di Vaio, A., Boccia, F., Landriani, L., Palladino, R., 2020. Artificial intelligence in the Agri- 105929.
food system: rethinking sustainable business models in the covid-19 scenario. Sus- Ishii, K., Matsuo, T., Takemura, Y., Sonoda, T., Nishida, Y., Yasukawa, S., Fujinaga, T., 2021.
tainability 12, 4851. https://doi.org/10.3390/su12124851. Tomato-harvesting-robot competition towards smart agriculture. Proceedings of In-
Ellafi, M.A., Deeks, L.K., Simmons, R.W., 2021. Application of artificial neural networks to ternational Conference on Artificial Life & Robotics (ICAROB2021), ALife Robotics ,
the design of subsurface drainage systems in libyan agricultural projects. J. Hydrol. pp. 1–5. https://alife-robotics.co.jp/ICAROB2021-PS1.pdf.
Reg. Stud. 35, 100832. https://doi.org/10.1016/j.ejrh.2021.100832. Jayne, T.S., Chamberlin, J., Headey, D.D., 2014. Land pressures, the evolution of farming
systems, and development strategies in africa: a synthesis. Food Policy 48, 1–17.
Emmi, L., Gonzalez-de Soto, M., Pajares, G., Gonzalez-de Santos, P., 2014. New trends in
https://doi.org/10.1016/j.foodpol.2014.05.014.
robotics for agriculture: integration and assessment of a real fleet of robots. Sci.
Jez, J.M., Topp, C.N., Yao, L., van de Zedde, R., Kowalchuk, G., 2021. Recent developments
World J. 2014. https://doi.org/10.1155/2014/404059.
and potential of robotics in plant eco-phenotyping. Emerg. Top. Life Sci. 5, 289–300.
Espejo-Garcia, B., Mylonas, N., Athanasakos, L., Fountas, S., 2020. Improving weeds identi-
https://doi.org/10.1042/ETLS20200275.
fication with a repository of agricultural pre-trained deep neural networks. Comput.
Jiang, H., Li, X., Safara, F., 2021. Iot-based agriculture: deep learning in detecting apple fruit
Electron. Agric. 175, 105593. https://doi.org/10.1016/j.compag.2020.105593.
diseases. Microprocess. Microsyst., 104321. https://doi.org/10.1016/j.micpro.2021.
Fahey, T., Pham, H., Gardi, A., Sabatini, R., Stefanelli, D., Goodwin, I., Lamb, D.W., 2021. Ac- 104321.
tive and passive electro-optical sensors for health assessment in food crops. Sensors
Jung, J., Maeda, M., Chang, A., Bhandari, M., Ashapure, A., Landivar-Bowles, J., 2021. The
21, 171. https://doi.org/10.3390/s21010171.
potential of remote sensing and artificial intelligence as tools to improve the resil-
Feng, Q., Zou, W., Fan, P., Zhang, C., Wang, X., 2018. Design and test of robotic harvesting ience of agriculture production systems. Curr. Opin. Biotechnol. 70, 15–22. https://
system for cherry tomato. Int. J. Agric. Biol. Eng. 11, 96–100. https://doi.org/10.25165/ doi.org/10.1016/j.copbio.2020.09.003.
j.ijabe.20181101.2853. Kakani, V., Nguyen, V.H., Kumar, B.P., Kim, H., Pasupuleti, V.R., 2020. A critical review on
Ferreras-Fernández, T., Merlo-Vega, J.A., Garca-Peñalvo, F.J., 2013. Impact of scientific con- computer vision and artificial intelligence in food industry. J. Agric. Food Res. 2,
tent in open access institutional repositories: a case study of the repository gredos. 100033. https://doi.org/10.1016/j.jafr.2020.100033.
Proceedings of the First International Conference on Technological Ecosystem for En- Kamilaris, A., Prenafeta-Boldú, F.X., 2018. Deep learning in agriculture: a survey. Comput.
hancing Multiculturality, pp. 357–363. https://doi.org/10.1145/2536536.2536590. Electron. Agric. 147, 70–90. https://doi.org/10.1016/j.compag.2018.02.016.
Franco, S., 2021. Assessing the environmental sustainability of local agricultural systems: Khan, T., Qiu, J., Qureshi, M.A.A., Iqbal, M.S., Mehmood, R., Hussain, W., 2020. Agricultural
how and why. Curr. Res. Environ. Sustain. 3, 100028. https://doi.org/10.1016/j.crsust. fruit prediction using deep neural networks. Proc. Comp. Sci. 174, 72–78. https://doi.
2021.100028. org/10.1016/j.procs.2020.06.058.
Fue, K.G., Porter, W.M., Barnes, E.M., Rains, G.C., 2020. An extensive review of mobile ag- Khare, D., Cherussery, S., Mohan, S., 2021. A novel design for an autonomous mobile agri-
ricultural robotics for field operations: focus on cotton harvesting. AgriEngineering 2, cultural fruit harvesting robot. IFToMM Symposium on Mechanism Design for Robot-
150–174. https://doi.org/10.3390/agriengineering2010010. ics. Springer, pp. 38–49. https://doi.org/10.1007/978-3-030-75271-2_5.
Gaffney, J., Bing, J., Byrne, P.F., Cassman, K.G., Ciampitti, I., Delmer, D., Habben, J., Lafitte, Khort, D., Kutyrev, A., Filippov, R., Semichev, S., 2021. Development control system robotic
H.R., Lidstrom, U.E., Porter, D.O., et al., 2019. Science-based intensive agriculture: sus- platform for horticulture. E3S Web of Conferences, EDP Sciences, p. 01024. https://
tainability, food security, and the role of technology. Global. Food Secur. 23, 236. doi.org/10.1051/e3sconf/202126201024.
https://doi.org/10.1016/j.gfs.2019.08.003. Kitchenham, B., Charters, S., 2007. Guidelines for performing Systematic Literature Re-
Gai, J., Xiang, L., Tang, L., 2021. Using a depth camera for crop row detection and mapping views in Software Engineering, Technical Report EBSE 2007-001. Keele University
for under-canopy navigation of agricultural robotic vehicle. Comput. Electron. Agric. and Durham University Joint Report.
188, 106301. https://doi.org/10.1016/j.compag.2021.106301. Kim, W.S., Lee, D.H., Kim, Y.J., Kim, T., Lee, W.S., Choi, C.H., 2021. Stereo-vision-based crop
Geissdoerfer, M., Savaget, P., Bocken, N.M., Hultink, E.J., 2017. The circular economy–a height estimation for agricultural robots. Comput. Electron. Agric. 181, 105937.
new sustainability paradigm? J. Clean. Prod. 143, 757–768. https://doi.org/10.1016/ https://doi.org/10.1016/j.compag.2020.105937.
j.jclepro.2016.12.048. Kiourt, C., Pavlidis, G., Markantonatou, S., 2020. Deep learning approaches in food recog-
Ghafar, A.S.A., Hajjaj, S.S.H., Gsangaya, K.R., Sultan, M.T.H., Mail, M.F., Hua, L.S., 2021. De- nition. Machine Learning Paradigms. Springer, pp. 83–108. https://doi.org/10.1007/
sign and development of a robot for spraying fertilizers and pesticides for agriculture. 978-3-030-49724-8_4.
Mater. Today: Proc. https://doi.org/10.1016/j.matpr.2021.03.174. Kounalakis, T., Triantafyllidis, G.A., Nalpantidis, L., 2019. Deep learning-based visual recog-
Giannakis, M., Spanaki, K., Dubey, R., 2019. A cloud-based supply chain management sys- nition of rumex for robotic precision farming. Comput. Electron. Agric. 165, 104973.
tem: effects on supply chain responsiveness. J. Enterp. Inf. Manag., 585–607. https:// https://doi.org/10.1016/j.compag.2019.104973.
doi.org/10.1108/JEIM-05-2018-0106. Kujawa, S., Niedbala, G., 2021. Artificial Neural Networks in Agriculture. https://doi.org/
Gonzalez-de Santos, P., Ribeiro, A., Fernandez-Quintanilla, C., Lopez-Granados, F., 10.3390/agriculture11060497.
Brandstoetter, M., Tomic, S., Pedrazzi, S., Peruzzi, A., Pajares, G., Kaplanis, G., et al., Kultongkham, A., Kumnon, S., Thintawornkul, T., Chanthasopeephan, T., 2021. The design
2017. Fleets of robots for environmentally-safe pest control in agriculture. Precis. of a force feedback soft gripper for tomato harvesting. J. Agric. Eng. 52. https://doi.
Agric. 18, 574–614. https://doi.org/10.1007/s11119-016-9476-3. org/10.4081/jae.2021.1090.
Kumar, P., Joshi, V., 2015. Applications of hydrological model swat on the upper water-
Grieve, B.D., Duckett, T., Collison, M., Boyd, L., West, J., Yin, H., Arvin, F., Pearson, S., 2019.
shed of river subarnarekha with special reference to model performance and its eval-
The challenges posed by global broadacre crops in delivering smart agri-robotic solu-
uation. J. Basic Appl. Eng. Res. 2, 1128–1134.
tions: a fundamental rethink is required. Glob. Food Secur. 23, 116–124. https://doi.
Kurniasih, D., Jasmi, K.A., Basiron, B., Huda, M., Maseleno, A., 2018. The uses of fuzzy logic
org/10.1016/j.gfs.2019.04.011.
method for finding agriculture and livestock value of potential village. Int. J. Eng.
Grimstad, L., From, P.J., 2017. The thorvald ii agricultural robotic system. Robotics 6, 24.
Technol. 7, 1091–1095. https://doi.org/10.14419/ijet.v7i3.12495.
https://doi.org/10.3390/robotics6040024.
Kwon, O.Y., Lee, M.H., Guan, C., Lee, S.W., 2019. Subject-independent brain–computer in-
Guillén, M.A., Llanes, A., Imbernón, B., Martnez-España, R., Bueno-Crespo, A., Cano, J.C., terfaces based on deep convolutional neural networks. IEEE Trans. Neural Netw.
Cecilia, J.M., 2021. Performance evaluation of edge-computing platforms for the pre- Learn. Syst. 31, 3839–3852. https://doi.org/10.1109/TNNLS.2019.2946869.
diction of low temperatures in agriculture using deep learning. J. Supercomput. 77,
Le, T., Gjevestad, J.G.O., From, P.J., 2019. Online 3D mapping and localization system for
818–840. https://doi.org/10.1007/s11227-020-03288-w.
agricultural robots. IFAC-PapersOnLine 52, 167–172. https://doi.org/10.1016/j.ifacol.
Gupta, M., Abdelsalam, M., Khorsandroo, S., Mittal, S., 2020. Security and privacy in smart 2019.12.516.
farming: challenges and opportunities. IEEE Access 8, 34564–34584. https://doi.org/ Lee, J., Nazki, H., Baek, J., Hong, Y., Lee, M., 2020. Artificial intelligence approach for tomato
10.1109/ACCESS.2020.2975142. detection and mass estimation in precision agriculture. Sustainability 12, 9138.
Han, S.Z., Pan, W.T., Zhou, Y.Y., Liu, Z.L., 2018. Construct the prediction model for China ag- https://doi.org/10.3390/su12219138.
ricultural output value based on the optimization neural network of fruit fly optimi- Li, J., Yin, J., Deng, L., 2021a. A robot vision navigation method using deep learning in edge
zation algorithm. Futur. Gener. Comput. Syst. 86, 663–669. https://doi.org/10.1016/j. computing environment. EURASIP Journal on Advances in Signal Processing 2021,
future.2018.04.058. pp. 1–20. https://doi.org/10.1186/s13634-021-00734-6.
Hespeler, S.C., Nemati, H., Dehghan-Niri, E., 2021. Non-destructive thermal imaging for Li, Q., Zeng, J., Lin, L., Zhang, J., Zhu, J., Yao, L., Wang, S., Du, J., Wu, Z., 2021b. Mid-infrared
object detection via advanced deep learning for robotic inspection and harvesting spectra feature extraction and visualization by convolutional neural network for
of chili peppers. Artific. Intellig. Agric. 5, 102–117. https://doi.org/10.1016/j.aiia. sugar adulteration identification of honey and real-world application. LWT 140,
2021.05.003. 110856. https://doi.org/10.1016/j.lwt.2021.110856.
Hossain, M.Z., Komatsuzaki, M., 2021. Weed management and economic analysis of a ro- Li, W., Wang, D., Li, M., Gao, Y., Wu, J., Yang, X., 2021c. Field detection of tiny pests from
botic lawnmower: a case study in a japanese pear orchard. Agriculture 11, 113. sticky trap images using deep learning in agricultural greenhouse. Comput. Electron.
https://doi.org/10.3390/agriculture11020113. Agric. 183, 106048. https://doi.org/10.1016/j.compag.2021.106048.
Huang, C.C., Chang, C.L., 2019. Design and implementation of bio-inspired snake bone- Libertn, M., Kvaček, J., Bek, J., Žárskỳ, V., Štorch, P., 2018. Sporophytes of polysporangiate
armed robot for agricultural irrigation application. IFAC-PapersOnLine 52, 98–101. land plants from the early silurian period may have been photosynthetically autono-
https://doi.org/10.1016/j.ifacol.2019.12.504. mous. Nat. Plants 4, 269–271. https://doi.org/10.1038/s41477-018-0140-y.

57
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Linaza, M.T., Posada, J., Bund, J., Eisert, P., Quartulli, M., Döllner, J., Pagani, A., Golaizola, I., Paliwal, N., Vanjani, P., Liu, J.W., Saini, S., Sharma, A., 2019. Image processing-based intel-
Barriguinha, A., Moysiadis, T., et al., 2021. Data-driven artificial intelligence applica- ligent robotic system for assistance of agricultural crops. Int. J. Soc. Human. Comp. 3,
tions for sustainable precision agriculture. Agronomy 11, 1227. https://doi.org/10. 191–204. https://doi.org/10.1504/IJSHC.2019.101602.
3390/agronomy11061227. Panarin, R.N., Khvorova, L.A., 2021. Software development for agricultural tillage robot
Liu, X., Zhu, X., Zhang, Q., Yang, T., Pan, Y., Sun, P., 2020. A remote sensing and artificial based on technologies of machine intelligence. International Conference on High-
neural network-based integrated agricultural drought index: index development Performance Computing Systems and Technologies in Scientific Research, Automa-
and applications. Catena 186, 104394. https://doi.org/10.1016/j.catena.2019.104394. tion of Control and Production. Springer, pp. 354–367. https://doi.org/10.1007/978-
Liu, L.W., Ma, X., Wang, Y.M., Lu, C.T., Lin, W.S., 2021. Using artificial intelligence algo- 3-030-94141-3_28.
rithms to predict rice (oryza sativa l.) growth rate for precision agriculture. Comput. Panda, S., 2014. Coordination of a socially responsible supply chain using revenue sharing
Electron. Agric. 187, 106286. https://doi.org/10.1016/j.compag.2021.106286. contract. Transp. Res. Part E: Logist. Transp. Rev. 67, 92–104. https://doi.org/10.1016/
Lytridis, C., Kaburlasos, V.G., Pachidis, T., Manios, M., Vrochidou, E., Kalampokas, T., j.tre.2014.04.002.
Chatzistamatis, S., 2021. An overview of cooperative robotics in agriculture. Agron- Peteinatos, G.G., Reichel, P., Karouta, J., Andújar, D., Gerhards, R., 2020. Weed identifica-
omy 11, 1818. https://doi.org/10.3390/agronomy11091818. tion in maize, sunflower, and potatoes with the aid of convolutional neural networks.
Magalhães, S.A., Castro, L., Moreira, G., Dos Santos, F.N., Cunha, M., Dias, J., Moreira, A.P., Remote Sens. 12, 4185. https://doi.org/10.3390/rs12244185.
2021. Evaluating the single-shot multibox detector and yolo deep learning models Porsch, M.R.M.H., Rasia, L.A., Thesing, N.J., Pedrali, P.C., Valdiero, A.C., 2019. Low cost ro-
for the detection of tomatoes in a greenhouse. Sensors 21, 3569. https://doi.org/10. botic manipulator for family agriculture. J. Agric. Stud. 7, 225–239. https://ideas.
3390/s21103569. repec.org/a/mth/jas888/v7y2019i2p225-239.html.
Magomadov, V., 2019. Deep learning and its role in smart agriculture. Journal of Physics: Pujari, D., Yakkundimath, R., Byadgi, A.S., 2016. Svm and ann based classification of plant
Conference Series. IOP Publishing, p. 044109. https://doi.org/10.1088/1742-6596/ diseases using feature reduction technique. IJIMAI 3, 6–14.
1399/4/044109. Raja, R., Nguyen, T.T., Slaughter, D.C., Fennimore, S.A., 2020. Real-time weed-crop classifi-
Mangla, S.K., Sharma, Y.K., Patil, P.P., Yadav, G., Xu, J., 2019. Logistics and distribution chal- cation and localisation technique for robotic weed control in lettuce. Biosyst. Eng.
lenges to managing operations for corporate sustainability: study on leading indian 192, 257–274. https://doi.org/10.1016/j.biosystemseng.2020.02.002.
diary organizations. J. Clean. Prod. 238, 117620. https://doi.org/10.1016/j.jclepro. Roshanianfard, A., Mengmeng, D., Nematzadeh, S., 2021. A 4-dof scara robotic arm for var-
2019.117620. ious farm applications: designing, kinematic modelling, and parameterization. Acta
Marinoudi, V., Sørensen, C.G., Pearson, S., Bochtis, D., 2019. Robotics and labour in agricul- Technol. Agric. 24, 61–66. https://doi.org/10.2478/ata-2021-0010.
ture. A context consideration. Biosyst. Eng. 184, 111–121. https://doi.org/10.1016/j. Ruigrok, T., van Henten, E., Booij, J., van Boheemen, K., Kootstra, G., 2020. Application-
biosystemseng.2019.06.013. specific evaluation of a weed-detection algorithm for plant-specific spraying. Sensors
McGuire, A.M., 2017. Agricultural science and organic farming: time to change our trajec- 20, 7262. https://doi.org/10.3390/s20247262.
tory. Agric. Environ. Lett. 2, 170024. https://doi.org/10.2134/ael2017.08.0024. Rysz, M., Mehta, S.S., 2021. A risk-averse optimization approach to human-robot collabo-
Mehta, S., Burks, T., 2016. Adaptive visual servo control of robotic harvesting systems. ration in robotic fruit harvesting. Comput. Electron. Agric. 182, 106018. https://doi.
IFAC-PapersOnLine 49, 287–292. https://doi.org/10.1016/j.ifacol.2016.10.053. org/10.1016/j.compag.2021.106018.
Mellor, J.W., 2017. Agricultural Development and Economic Transformation: Promoting Santin, F., Da Silva, R., Grzybowski, J., 2016. Artificial neural network ensembles and the
Growth with Poverty Reduction. Springer. https://doi.org/10.1007/978-3-319- design of performance-oriented riparian buffer strips for the filtering of nitrogen in
65259-7. agricultural catchments. Ecol. Eng. 94, 493–502. https://doi.org/10.1016/j.ecoleng.
Mendes, J.M., dos Santos, F.N., Ferraz, N.A., do Couto, P.M., dos Santos, R.M., 2019. Localiza- 2016.06.008.
tion based on natural features detector for steep slope vineyards. J. Intell. Robot. Syst. Santos, L., Santos, F.N., Oliveira, P.M., Shinde, P., 2019. Deep learning applications in agri-
93, 433–446. https://doi.org/10.1007/s10846-017-0770-8. culture: a short review. Iberian Robotics Conference. Springer, pp. 139–151. https://
Midtiby, H.S., Åstrand, B., Jørgensen, O., Jørgensen, R.N., 2016. Upper limit for context– doi.org/10.1007/978-3-030-35990-4_12.
based crop classification in robotic weeding applications. Biosyst. Eng. 146, Santos, L., Santos, F., Mendes, J., Costa, P., Lima, J., Reis, R., Shinde, P., 2020. Path planning
183–192. https://doi.org/10.1016/j.biosystemseng.2016.01.012. aware of robot’s center of mass for steep slope vineyards. Robotica 38, 684–698.
Mohamed, E., Ali, A., El-Shirbeny, M., Abutaleb, K., Shaddad, S.M., 2020. Mapping soil https://doi.org/10.1017/S0263574719000961.
moisture and their correlation with crop pattern using remotely sensed data in arid Sayer, J., Cassman, K.G., 2013. Agricultural innovation to protect the environment. Proc.
region. Egypt. J. Remote Sens. Space Sci. 23, 347–353. https://doi.org/10.1016/j.ejrs. Natl. Acad. Sci. 110, 8345–8348. https://doi.org/10.1073/pnas.1208054110.
2019.04.003. Schor, J.B., Attwood-Charles, W., 2017. The “sharing” economy: labor, inequality, and so-
Mohammed, I.I., Jassim, A.R.A.L., 2021. Design and testing of an agricultural robot to oper- cial connection on for-profit platforms. Sociol. Compass 11, e12493. https://doi.org/
ate a combined seeding machine. Ann. Roman. Soc. Cell Biol. 25, 92–106. https:// 10.1111/soc4.12493.
annalsofrscb.ro/index.php/journal/article/view/9180. Senocak, A.A., Goren, H.G., 2021. Forecasting the biomass-based energy potential using ar-
Mohapatra, A.G., Lenka, S.K., 2016. Neural network pattern classification and weather de- tificial intelligence and geographic information systems: a case study. Eng. Sci.
pendent fuzzy logic model for irrigation control in WSN based precision agriculture. Technol. Int. J. 26, 100992. https://doi.org/10.1016/j.jestch.2021.04.011.
Proc. Comp. Sci. 78, 499–506. https://doi.org/10.1016/j.procs.2016.02.094. Seo, Y., Umeda, S., 2021. Evaluating farm management performance by the choice of pest-
Monteiro, A.L., de Freitas Souza, M., Lins, H.A., da Silva Teófilo, T.M., Júnior, A.P.B., Silva, control sprayers in rice farming in Japan. Sustainability 13, 2618. https://doi.org/10.
D.V., Mendonça, V., 2021. A new alternative to determine weed control in agricultural 3390/su13052618.
systems based on artificial neural networks (anns). Field Crop Res. 263, 108075. Sharma, A.K., Bisen, U.K., 2013. Taxonomic documentation of insect pest fauna of vegeta-
https://doi.org/10.1016/j.fcr.2021.108075. ble ecosystem collected in light trap. Int. J. Environ. Sci. Develop. Monitor. (IJESDM) 4,
Moya-Rico, J., Molina, A.E., Belmonte, J.F., Tendero, J.C., Almendros-Ibanez, J.A., 2019. Char- 4–10 URL https://www.ripublication.com/ijesdmspl/ijesdmv4n3_01.pdf.
acterization of a triple concentric-tube heat exchanger with corrugated tubes using Sharma, A., Jain, A., Gupta, P., Chowdary, V., 2020. Machine learning applications for pre-
artificial neural networks (ann). Appl. Therm. Eng. 147, 1036–1046. https://doi.org/ cision agriculture: a comprehensive review. IEEE Access 9, 4843–4873. https://doi.
10.1016/j.applthermaleng.2018.10.136. org/10.1109/ACCESS.2020.3048415.
Navarro, E., Costa, N., Pereira, A., 2020. A systematic review of iot solutions for smart farm- Sharma, A., Kumar, H., Mittal, K., Kauhsal, S., Kaushal, M., Gupta, D., Narula, A., 2021. Iot
ing. Sensors 20, 4231. https://doi.org/10.3390/s20154231. and deep learning-inspired multi-model framework for monitoring active fire loca-
Navas, E., Fernandez, R., Sepúlveda, D., Armada, M., Gonzalez-de Santos, P., 2020. A design tions in agricultural activities. Comput. Electr. Eng. 93, 107216. https://doi.org/10.
criterion based on shear energy consumption for robotic harvesting tools. Agronomy 1016/j.compeleceng.2021.107216.
10, 734. https://doi.org/10.3390/agronomy10050734. Singh, A., Imtiyaz, M., Isaac, R., Denis, D., 2012. Comparison of soil and water assessment
Ngugi, L.C., Abelwahab, M., Abo-Zahhad, M., 2021. Recent advances in image processing tool (swat) and multilayer perceptron (mlp) artificial neural network for predicting
techniques for automated leaf pest and disease recognition–a review. Inform. Proc. sediment yield in the nagwa agricultural watershed in Jharkhand, India. Agric.
Agric. 8, 27–51. https://doi.org/10.1016/j.inpa.2020.04.004. Water Manag. 104, 113–120. https://doi.org/10.1016/j.agwat.2011.12.005.
Nguyen, T.H., Nong, D., Paustian, K., 2019. Surrogate-based multi-objective optimization Song, Z., Zhou, Z., Wang, W., Gao, F., Fu, L., Li, R., Cui, Y., 2021. Canopy segmentation and
of management options for agricultural landscapes using artificial neural networks. wire reconstruction for kiwifruit robotic harvesting. Comput. Electron. Agric. 181,
Ecol. Model. 400, 1–13. https://doi.org/10.1016/j.ecolmodel.2019.02.018. 105933. https://doi.org/10.1016/j.compag.2020.105933.
Nguyen, T.T., Hoang, T.D., Pham, M.T., Vu, T.T., Nguyen, T.H., Huynh, Q.T., Jo, J., 2020. Mon- Spanaki, K., Karafili, E., Sivarajah, U., Despoudi, S., Irani, Z., 2021. Artificial intelligence and
itoring agriculture areas with satellite images and deep learning. Appl. Soft Comput. food security: swarm intelligence of agritech drones for smart agrifood operations.
95, 106565. https://doi.org/10.1016/j.asoc.2020.106565. Prod. Plan. Control, 1–19 https://doi.org/10.1080/09537287.2021.1882688.
Oliveira, L.F., Moreira, A.P., Silva, M.F., 2021. Advances in agriculture robotics: a state-of- Sudars, K., Jasko, J., Namatevs, I., Ozola, L., Badaukis, N., 2020. Dataset of annotated food
the-art review and challenges ahead. Robotics 10, 52. https://doi.org/10.3390/robot- crops and weed images for robotic computer vision control. Data Brief 31, 105833.
ics10020052. https://doi.org/10.1016/j.dib.2020.105833.
Özlüoymak, Ö.B., Bolat, A., 2020. Development and assessment of a novel imaging soft- Taghavifar, H., Mardani, A., Hosseinloo, A.H., 2015. Appraisal of artificial neural network-
ware for optimizing the spray parameters on water-sensitive papers. Comput. Elec- genetic algorithm based model for prediction of the power provided by the agricul-
tron. Agric. 168, 105104. https://doi.org/10.1016/j.compag.2019.105104. tural tractors. Energy 93, 1704–1710. https://doi.org/10.1016/j.energy.2015.10.066.
Özlüoymak, Ö.B., Bolat, A., Bayat, A., Güzel, E., 2019. Design, development, and evaluation Thomopoulos, V., Bitas, D., Papastavros, K.N., Tsipianitis, D., Kavga, A., 2021. Development
of a target oriented weed control system using machine vision. Turk. J. Agric. For. 43, of an integrated iot-based greenhouse control three-device robotic system. Agron-
164–173. https://doi.org/10.3906/tar-1803-8. omy 11, 405. https://doi.org/10.3390/agronomy11020405.
Padilla, J.E., Hudson, A., 2019. United nations development programme (undp) perspec- Tian, Z., Wang, J.W., Li, J., Han, B., 2021. Designing future crops: challenges and strategies
tives on asian large marine ecosystems. Deep-Sea Res. II Top. Stud. Oceanogr. 163, for sustainable agriculture. Plant J. 105, 1165–1178. https://doi.org/10.1111/tpj.
127–129. https://doi.org/10.1016/j.dsr2.2019.05.011. 15107.

58
V. Sachithra and L.D.C.S. Subhashini Artificial Intelligence in Agriculture 8 (2023) 46–59

Tilman, D., Balzer, C., Hill, J., Befort, B.L., 2011. Global food demand and the sustainable in- Yorozu, A., Ishigami, G., Takahashi, M., 2021. Human-following control in furrow for agri-
tensification of agriculture. Proc. Natl. Acad. Sci. 108, 20260–20264. https://doi.org/ cultural support robot. International Conference on Intelligent Autonomous Systems.
10.1073/pnas.1116437108. Springer, pp. 155–164. https://doi.org/10.1007/978-3-030-95892-3_12.
Tittonell, P., Giller, K.E., 2013. When yield gaps are poverty traps: the paradigm of ecolog- Zangina, U., Buyamin, S., Abidin, M., Mahmud, M., 2021. Agricultural rout planning with
ical intensification in african smallholder agriculture. Field Crop Res. 143, 76–90. variable rate pesticide application in a greenhouse environment. Alexandr. Eng. J.
https://doi.org/10.1016/j.fcr.2012.10.007. 60, 3007–3020. https://doi.org/10.1016/j.aej.2021.01.010.
Traldi, R., 2021. Progress and pitfalls: a systematic review of the evidence for agricultural Zapotezny-Anderson, P., Lehnert, C., 2019. Towards active robotic vision in agriculture: a
sustainability standards. Ecol. Indic. 125, 107490. https://doi.org/10.1016/j.ecolind. deep learning approach to visual servoing in occluded and unstructured protected
2021.107490. cropping environments. IFAC-PapersOnLine 52, 120–125. https://doi.org/10.1016/j.
Ünal, I., Topakci, M., Çanakci, M., Karayel, D., Yilmaz, E., Kabas, Ö., 2021. Development of a ifacol.2019.12.508.
mobile robot-based combined sensor platform to determine the corre-lation be- Zargar, G., Tanha, A.A., Parizad, A., Amouri, M., Bagheri, H., 2020. Reservoir rock properties
tween soil penetration resistance and electrical conductivity. Turk. J. Agric. For. 45, estimation based on conventional and nmr log data using ann-cuckoo: a case study in
365–379. https://doi.org/10.3906/tar-2007-43. one of super fields in Iran southwest. Petroleum 6, 304–310. https://doi.org/10.1016/
Utstumo, T., Urdal, F., Brevik, A., Dørum, J., Netland, J., Overskeid, Ø., Berge, T.W., Gravdahl, j.petlm.2019.12.002.
J.T., 2018. Robotic in-row weed control in vegetables. Comput. Electron. Agric. 154, Zhang, C., Noguchi, N., 2017. Development of a multi-robot tractor system for agriculture
36–45. https://doi.org/10.1016/j.compag.2018.08.043. field work. Comput. Electron. Agric. 142, 79–90. https://doi.org/10.1016/j.compag.
Väljaots, E., Lehiste, H., Kiik, M., Leemet, T., 2018. Soil sampling automation using mobile 2017.08.017.
robotic platform. J. Agron. Res. 16 (3), 917–922. https://dspace.emu.ee/bitstream/ Zhang, B., Xie, Y., Zhou, J., Wang, K., Zhang, Z., 2020. State-of-the-art robotic grippers,
handle/10492/4000/Vol16No3_30.pdf?sequence=4isAllowed=y. grasping and control strategies, as well as their applications in agricultural robots: a
Villa-Henriksen, A., Edwards, G.T.C., Green, O., Sørensen, C.A.G., 2021. Evaluation of grain review. Comput. Electron. Agric. 177, 105694. https://doi.org/10.1016/j.compag.
quality-based simulated selective harvest performed by an autonomous agricultural 2020.105694.
robot. Agronomy 11, 1728. https://doi.org/10.3390/agronomy11091728. Zhang, L., Huang, Z., Liu, W., Guo, Z., Zhang, Z., 2021. Weather radar echo prediction
Vincent, D.R., Deepa, N., Elavarasan, D., Srinivasan, K., Chauhdary, S.H., Iwendi, C., 2019. method based on convolution neural network and long short-term memory net-
Sensors driven ai-based agriculture recommendation model for assessing land suit- works for sustainable e-agriculture. J. Clean. Prod. 298, 126776. https://doi.org/10.
ability. Sensors 19, 3667. https://doi.org/10.3390/s19173667. 1016/j.jclepro.2021.126776.
Wheeler, T., Von Braun, J., 2013. Climate change impacts on global food security. Science Zhou, Z., Majeed, Y., Naranjo, G.D., Gambacorta, E.M., 2021. Assessment for crop water
341, 508–513. https://doi.org/10.1126/science.1239402. stress with infrared thermal imagery in precision agriculture: a review and future
Williams, H.A., Jones, M.H., Nejati, M., Seabright, M.J., Bell, J., Penhall, N.D., Barnett, J.J., prospects for deep learning applications. Comput. Electron. Agric. 182, 106019.
Duke, M.D., Scarfe, A.J., Ahn, H.S., et al., 2019. Robotic kiwifruit harvesting using ma- https://doi.org/10.1016/j.compag.2021.106019.
chine vision, convolutional neural networks, and robotic arms. Biosyst. Eng. 181, Zhu, N., Liu, X., Liu, Z., Hu, K., Wang, Y., Tan, J., Huang, M., Zhu, Q., Ji, X., Jiang, Y., et al., 2018.
140–156. https://doi.org/10.1016/j.biosystemseng.2019.03.007. Deep learning for smart agriculture: concepts, tools, applications, and opportunities.
Xue, W., Hu, X., Wei, Z., Mei, X., Chen, X., Xu, Y., 2019. A fast and easy method for Int. J. Agric. Biol. Eng. 11, 32–44. https://doi.org/10.25165/j.ijabe.20181104.4475.
predicting agricultural waste compost maturity by image-based deep learning. Zujevs, A., Osadcuks, V., Ahrendt, P., 2015. Trends in robotic sensor technologies for fruit
Bioresour. Technol. 290, 121761. https://doi.org/10.1016/j.biortech.2019.121761. harvesting: 2010-2015. Proc. Comp. Sci. 77, 227–233. https://doi.org/10.1016/j.procs.
Yamaç, S.S., 2021. Artificial intelligence methods reliably predict crop evapotranspiration 2015.12.378.
with different combinations of meteorological data for sugar beet in a semiarid area.
Agric. Water Manag. 254, 106968. https://doi.org/10.1016/j.agwat.2021.106968.

59