Technical Note
A systematic literature review on machine learning applications for sustainable agriculture supply chain performance

https://doi.org/10.1016/j.cor.2020.104926Get rights and content

Highlights

Abstract

Agriculture plays an important role in sustaining all human activities. Major challenges such as overpopulation, competition for resources poses a threat to the food security of the planet. In order to tackle the ever-increasing complex problems in agricultural production systems, advancements in smart farming and precision agriculture offers important tools to address agricultural sustainability challenges. Data analytics hold the key to ensure future food security, food safety, and ecological sustainability. Disruptive information and communication technologies such as machine learning, big data analytics, cloud computing, and blockchain can address several problems such as productivity and yield improvement, water conservation, ensuring soil and plant health, and enhance environmental stewardship. The current study presents a systematic review of machine learning (ML) applications in agricultural supply chains (ASCs). Ninety three research papers were reviewed based on the applications of different ML algorithms in different phases of the ASCs. The study highlights how ASCs can benefit from ML techniques and lead to ASC sustainability. Based on the study findings an ML applications framework for sustainable ASC is proposed. The framework identifies the role of ML algorithms in providing real-time analytic insights for pro-active data-driven decision-making in the ASCs and provides the researchers, practitioners, and policymakers with guidelines on the successful management of ASCs for improved agricultural productivity and sustainability.

Introduction

Agriculture sustainability is the key to ensure food security and hunger eradication for the ever-growing population. It is estimated that global food production must be increased by 60–110% to feed 9–10 billion of the population by 2050 (Tilman et al., 2011; Pardey et al., 2014; Rockström et al., 2017). It is therefore required to have a strategic shift from the current paradigm of enhanced agricultural productivity to agricultural sustainability (Rockström et al., 2017). Sustainable agriculture practices not only focus on enriching agricultural productivity but also help to reduce harmful environmental impacts (Kuyper and Struik, 2014; Godfray and Garnett, 2014; Cobuloglu and Büyüktahtakın, 2015; Adnan et al., 2018). The sustainable Agriculture Supply Chains (ASCs) are knowledge-intensive and are based on information, skills, technologies, and attitudes of the supply chain partners (El Bilali and Allahyari, 2018). Knowledge transfer encourages farmers to enhance their decision to adopt sustainable agriculture practices (SAP) (Adnan et al., 2018).

Strothkämper (2016) claims that the ASCs are facing tremendous pressure to increase the farming efficiency, which is driven by the depleting rate of water and fossil fuels, shrinking availability of arable land and the increasing demand by the consumers for more transparent and sustainable food chains (Tian, 2016, Duman et al., 2017). The need for the ASCs to respond to the increasing demand and supply gaps, as well as market price fluctuations, is also identified as critical drivers of farming efficiency (Sharma et al., 2018; Patidar et al., 2018). Further, recent studies covering the sustainable aspects of inventory and transportation management concerning the perishable items may help us to understand the complexities involved in achieving sustainable ASCs. The digital technologies that include the internet of things (IoT), mobile technologies and devices, data analytics, artificial intelligence (AI), digitally delivered services, and other applications are influencing the ASCs (Kamilaris and Prenafeta-Boldú, 2018). Numerous examples demonstrate the use of digital technologies at different stages of ASC such as automation of farm machinery resulting in reduced labour input, use of sensors and remote satellite data for improved monitoring of crops, land, and water, IoT and RFID for agriculture product traceability (OECD, 2019)

As a result of going digital, a large amount of data is getting generated in the supply chains, which is useless unless it is organized, understood, and meaningful insights are gained using appropriate data analysis tools (Russo et al., 2015, Dubey et al., 2015). AI or cognitive-based technologies is the most transformative and impactful advanced analytics tool that can be used by the organizations for supply chain decision making (Liakos et al., 2018). AI helps computers interact, reason, and learn like human beings to enable them to perform a wide variety of cognitive tasks, usually requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages and demonstrating an ability to move and manipulate objects accordingly. Intelligent systems use a combination of big data analytics, cloud computing, machine-to-machine communication, and the IoT to operate and learn (OECD, 2017). Machine and deep learning algorithms, the subsets of AI, is widely used in combination with location intelligence technologies in ASC to identify hidden patterns in the data (Elavarasan et al., 2018). The study by Kazemi (2019) published in Forbes suggests ASC practitioners consider AI and advanced analytics as strategic investments because of the accelerating digital transformation in ASC and need to develop a competitive advantage. Patidar et al. (2018) described how information technology (IT) is impacting the expectations of farmers as well as customers and accordingly, their demand is changing.

In the past, few review studies were conducted on AI and ML applications for improving the supply chain performance, as mentioned in Table 1. These studies have focused on ML applications in the supply chain management (Min, 2010) covering specific aspects like supply chain risk management (Baryannis et al. 2019) or sectors (Ngai et al., 2014; Konovalenko and Ludwig, 2019). However, to the best of our knowledge, no such studies are conducted to review the present status of AI/ML applications in ASC management. The increasing amount of data captured by emerging technologies offer the ASCs new abilities to predict changes and identify opportunities (Kamble et al., 2020; Kamble et al., 2019a). The practitioners must be equipped with the latest knowledge to ensure that significant insights are derived from the collected data. Extensive testing and validation of emerging ML applications in ASC will be critical as agriculture is impacted by environmental factors that cannot be controlled, unlike other industries where risk is more comfortable to model and predict (Sennaar, 2019). Therefore, in this study, we present a systematic literature review (SLR) of 93 papers on ML applications in developing sustainable ASC. It is anticipated that the agricultural sector will continue to see the increasing adoption of ML in future and the results of this study will guide the researchers and practitioners to understand the present status of ML applications in ASC, which will help them to understand how adoption of ML will support the ASC to optimize farming practices to increase yields, crop quality and incomes in a sustainable manner.

The remaining of the paper is organized as follows: Section 2 presents a brief on sustainable ASC and ML algorithms. Section 3 presents the SLR methodology adopted in this study. Section 4 discusses the results of the SLR. An ML-ASC framework and implications based on the findings of the study are discussed in Section 5. The conclusions and limitations of the study are presented in Section 6.

Section snippets

Agriculture supply chain

ASCs are like the fast-moving consumer goods (FMCG) supply chains in many ways but differ in terms of raw material procurement and the final product. The raw materials are procured from the fields, and the product is made for consumption by humans or animals. As seen in Fig. 1, the ASC includes several operations such as pre-production, production, storage, processing, retail, and distribution before the final product reaches the end consumers (Borodin et al., 2016).

A typical ASC includes

Review methodology

In this study, we utilized an SLR with a specific focus on reviewing the published research work systematically and attain an unbiased and objective summary of the current state and future potential of ML applications in ASC. SLR adopts an approach that is scientific and replicable (Cook et al. 1997) for evaluating and interpreting all the available research relevant to a question, topic, or phenomenon of interest (Booth et al. 2012). SLRs help in developing useful insights based on the

Review findings and discussions

In this section, we analyze and present the findings of SLR in four clusters. These clusters represent the use of ML in the different ASC phases and include; pre-production, production, processing, and distribution (Aramyan et al., 2006; Ahumada and Villalobos, 2009). The use of ML in each cluster serves specific applications for improving ASC efficiency. It is observed from the SLR that ML is applied in the pre-production phase (Cluster 1) for the prediction of crop yield, soil properties, and

ML-ASC performance framework

The review findings indicate that ML has a vast potential for applications in the different ASC phases. The data generated from different sources in the ASC is used for making predictions and classification using different ML algorithms. The findings indicate that ML-driven technologies support improving the overall efficiency of ASC and address the various challenges faced by the industry, such as crop yield, soil health, and disease management. The potential benefits lead to an improvement in

Summary

The present study is based on an SLR to investigate the current state of research on machine learning (ML) applications in ASC. The SLR was performed on 93 research articles, which were categorized using different ML algorithms across different ASC phases. The study finds that all three ML algorithms, that is, supervised, unsupervised, and reinforcement learning is used to develop sustainable ASCs. The main contribution of the study is the ML-ASC performance application framework (as shown in

References (206)

  • V. Borodin et al.

    Handling uncertainty in agricultural supply chain management: a state of the art

    Eur. J. Oper. Res.

    (2016)
  • N. Chirinda et al.

    Emissions of nitrous oxide from arable organic and conventional cropping systems on two soil types

    Agr. Ecosyst. Environ.

    (2010)
  • A. Chlingaryan et al.

    Machine learning approaches for crop yield prediction and nitrogen status estimation in precision agriculture: a review

    Comput. Electron.n Agric.

    (2018)
  • H.I. Cobuloglu et al.

    A stochastic multi-criteria decision analysis for sustainable biomass crop selection

    Expert Syst. Appl.

    (2015)
  • E.J. Coopersmith et al.

    Machine learning assessments of soil drying for agricultural planning

    Comput. Electron. Agric.

    (2014)
  • S. Cramer et al.

    An extensive evaluation of seven machine learning methods for rainfall prediction in weather derivatives

    Expert Syst. Appl.

    (2017)
  • M. Craninx et al.

    Artificial neural network models of the rumen fermentation pattern in dairy cattle

    Computers and Electronics in Agriculture

    (2008)
  • P. Devapriya et al.

    Integrated production and distribution scheduling with a perishable product

    Eur. J. Oper. Res.

    (2017)
  • B. Drury et al.

    A survey of the applications of Bayesian networks in agriculture

    Eng. Appl. Artif. Intell.

    (2017)
  • C.J. Du et al.

    Learning techniques used in computer vision for food quality evaluation: a review

    J. Food Eng.

    (2006)
  • G.M. Duman et al.

    A holistic approach for performance evaluation using quantitative and qualitative data: a food industry case study

    Expert Syst. Appl.

    (2017)
  • H. El Bilali et al.

    Transition towards sustainability in agriculture and food systems: role of information and communication technologies

    Inf. Process. Agric.

    (2018)
  • D. Elavarasan et al.

    Forecasting yield by integrating agrarian factors and machine learning models: a survey

    Comput. Electron. Agric.

    (2018)
  • B. Espejo-Garcia et al.

    Machine learning for automatic rule classification of agricultural regulations: a case study in Spain

    Comput. Electron. Agric.

    (2018)
  • C. Folberth et al.

    Spatio-temporal downscaling of gridded crop model yield estimates based on machine learning

    Agr. Forest Meteorol.

    (2019)
  • A. Goap et al.

    An IoT based smart irrigation management system using machine learning and open source technologies

    Comput. Electron. Agric.

    (2018)
  • K. Govindan et al.

    Two-echelon multiple-vehicle location–routing problem with time windows for optimization of sustainable supply chain network of perishable food

    Int. J. Prod. Econ.

    (2014)
  • A. Haghverdi et al.

    Prediction of cotton lint yield from phenology of crop indices using artificial neural networks

    Comput. Electron. Agric.

    (2018)
  • S.S. Kamble et al.

    Achieving sustainable performance in a data-driven agriculture supply chain: a review for research and applications

    Int. J. Prod. Econ.

    (2020)
  • S.S. Kamble et al.

    Sustainable industry 4.0 framework: a systematic literature review identifying the current trends and future perspectives

    Process Saf. Environ. Protect.

    (2018)
  • S.S. Kamble et al.

    Modeling the internet of things adoption barriers in food retail supply chains

    J. Retail. Consum. Serv.

    (2019)
  • A. Kamilaris et al.

    Deep learning in agriculture: a survey

    Comput. Electron. Agric.

    (2018)
  • A. Kamilaris et al.

    A review on the practice of big data analysis in agriculture

    Comput. Electron. Agric.

    (2017)
  • Y. Kaneda et al.

    Multi-modal sliding window-based support vector regression for predicting plant water stress

    Knowl. Based Syst.

    (2017)
  • S. Khanal et al.

    Integration of high resolution remotely sensed data and machine learning techniques for spatial prediction of soil properties and corn yield

    Comput. Electron. Agric.

    (2018)
  • B. Khoshnevisan et al.

    Decreasing environmental impacts of cropping systems using life cycle assessment (LCA) and multi-objective genetic algorithm

    J. Clean. Prod.

    (2015)
  • I. Konovalenko et al.

    Event processing in supply chain management–The status quo and research outlook

    Comput. Ind.

    (2019)
  • L. Kouadio et al.

    Artificial intelligence approach for the prediction of Robusta coffee yield using soil fertility properties

    Comput. Electron. Agric.

    (2018)
  • T. Krisztin

    Semi-parametric spatial autoregressive models in freight generation modeling

    Transp. Res. Part E

    (2018)
  • M. Kummu et al.

    Lost food, wasted resources: global food supply chain losses and their impacts on freshwater, cropland, and fertiliser use

    Sci. Total Environ.

    (2012)
  • T.W. Kuyper et al.

    Epilogue: global food security, rhetoric, and the sustainable intensification debate

    Curr. Opin. Environ. Sustain.

    (2014)
  • W.S. Lee et al.

    Sensing technologies for precision specialty crop production

    Comput. Electron. Agric.

    (2010)
  • P. Abbal et al.

    A decision support system for vine growers based on a Bayesian network

    J. Agric. Biol. Environ. Stat.

    (2016)
  • N. Adnan et al.

    The effects of knowledge transfer on farmers decision making toward sustainable agriculture practices: in view of green fertilizer technology

    World J. Sci., Technol. Sustain. Dev.

    (2018)
  • Ahn, H.S., Dayoub, F., Popovic, M., MacDonald, B., Siegwart, R. and Sa, I., 2018. An overview of perception methods for...
  • L. Aramyan et al.

    Performance indicators in agri-food production chains

    Quantify. Agri-Food Supply Chain

    (2006)
  • D. Armstrong et al.

    The place of inter-rater reliability in qualitative research: an empirical study

    Sociology

    (1997)
  • F. Balducci et al.

    Machine learning applications on agricultural datasets for smart farm enhancement

    Machines

    (2018)
  • G. Baryannis et al.

    Supply chain risk management and artificial intelligence: state of the art and future research directions

    Int. J. Prod. Res.

    (2019)
  • J.A. Beaman et al.

    Food distribution channel overview: a guide for new manufacturers

    (2006)
  • Cited by (0)

    1.

    Research Scholar (Fellow).

    2.

    Associate Professor.

    3.

    Dean and Professor.

    4.

    Professor.

    5.

    Post-Doctoral Research Fellow.

    View full text