Agro 4.0: A data science-based information system for sustainable agroecosystem management

https://doi.org/10.1016/j.simpat.2020.102068Get rights and content

Abstract

One of the solutions for handling and treating the diverse data related to the sustainability of an agroecosystem is the use of Information Systems and Internet of Things. In this work, we adopt a methodology called Indicators of Sustainability in Agroecosystems (Indicadores de Sustentabilidade em Agroecossistemas – ISA), implement an information system based on Internet of Things and apply Data Science and simulation techniques over the gathered data, from 100 real rural properties. As a result, we have developed a set of tools for data collection, processing, visualization, simulation and analysis of the sustainability of a rural property or region, following the ISA methodology. Two experiments were applied on the dataset collected by the tools: environmental change scenarios simulations on targeted agroecosystems to predict how they affect two ISA scores (Soil Fertility and Water Quality) of involved agroecosystems; Evaluation of Feature Selection models searching for subsets of features good enough to predict the two ISA scores for the dataset with a smaller amount of data necessary. We have that with only 7 of the 21 Indicators present in ISA we can identify the level of sustainability in more than 90% of cases, allowing for a new discussion about shrinking the amount of data needed for the computation of ISA, or remodeling the final computation of the Sustainability Index so other Indicators can be more expressive. Users of the solutions developed in this work can identify best practices for sustainability in participating agroecosystems.

Introduction

Sustainability is when one can work on the present development without compromising the development of future generations [1]. The idea of sustainable development became widely known after the United Nations Conference on Environment and Development in Rio de Janeiro in 1992 (ECO-92). Since then, there have been variations in the definition of sustainability [2], [3], [4], [5], but they all converge to a definition where sustainability implies a medium- and long-run profitability, as well as agricultural practices with sustainable environmental impacts. Public awareness of the negative impacts of human activity on our environment is at an all times high, and, according to specialists’ predictions, no more time can be wasted [6].

Achieving the growing challenges of the next decade [7] is a complex task due to a large number of variables. In addition to a solid model for measuring sustainability, relevant data needs to be well structured for retrieval, storage and analysis. There are also other problems related to understanding the factors that affect sustainability. Therefore, there is a dire need for a system able to: a) characterize, visualize, and analyze collected data; and b) implement smart strategies that can measure sustainability in agroecosystems.

As a solution to this problem, we have designed, developed, and evaluated a sustainability management system in agroecosystems based on data science and internet of things. The system, dubbed Agro 4.01 [8], provides tools for the collection, storage, analysis, visualization and scenarios simulations of sustainability-related information of rural properties. Agro 4.0  enables, for instance, different properties to be compared regarding their Sustainability Index. Besides, the system can pinpoint agro-ecosystems with critical levels of sustainability and then suggest managers measures to reverse the situation. It is worth saying that Agro 4.0 is based on the Agroecosystems Sustainability Index (ISA) [9] methodology.

Agro 4.0 is inserted in a multidisciplinary and multifaceted context. First of all, it is an Green Information System that supports Decision Making with data, information extraction and visualization. The system facilitates the work of groups and entities that seek to improve the sustainability of properties of a given profile. This profile can be, for instance, for the type of product produced. The system has different interactive interfaces (for the browser, tablets and desktop), and asks for and accepts data potentially generated by other sensors (cameras, soil and water sensors, drones, etc). The system makes use of Data Science, which enables managers of projects to perform more complex analysis over a set of rural properties. Agroecosystems and Sustainability are at the core of the ISA Methodology, that Agro 4.0  implements.

This work has three main contributions: (i) the identification of which indicators of the ISA Model are more relevant or expressive for the Sustainability of a rural property, opening a discussion about the amount of data that ISA requires and the possibility to reduce the input (as it is right now, there are hundreds of fields a technician has to fill in order to obtain the final scores, and some require analysis of samples in laboratories and portable test kits); (ii) A tool to simulate environmental changes scenarios and their impact on the sustainability score and other indicators of the ISA methodology on participating agroecosystems; (iii) a data science-based information system for sustainability management of agroecosystems that allows to:

  • 1.

    collect, to structure and validate data about the sustainability of agroecosystems using the ISA Methodology;

  • 2.

    manage information about sustainability and support decision in agroecosystems;

  • 3.

    identify and characterize the most relevant factors for sustainability;

  • 4.

    perform visualization and analysis over aggregate data in a user-friendly way;

  • 5.

    simulate environmental changes and measure their impact on ISA indicators and sustainability indexes.

Besides, we have validated Agro 4.0  by using data from one hundred rural properties in the state of Minas Gerais, Brazil. Minas Gerais was Brazil’s greatest producer of both coffee and milk in 2018. Brazil was the world’s greatest producer and exporter of coffee in 2018 and in previous years as well. Brazil was also the worlds’ greatest exporter of both beef and chicken in the same year. There were more cattle heads than people in Brazil in the year 2018, according to FAS/USDA.

The rest of this work is organized as follows. In Section 2 we present the theoretical basis of our work describing concepts of agroecosystems sustainability, and data science. In Section 3 we discuss the related work, concepts of agroecosystems, sustainability and data science. In Section 4 we present our platform and its architecture. In Section 5 we describe a case study based on rural properties located in the countryside of Brazil. Finally, in Section 7, we conclude our work.

Section snippets

Fundamentals

In this section, we describe the theoretical fundamentals needed for the comprehension of this paper. Section 2.1 describes the main concepts related to agroecosystems sustainability, focusing on the ISA model and methodology.

Related works

The article [34] made a review of the literature for big data applications in farming and agriculture. Thirty-four articles were analysed for the tools they used and the problems they tackled. The authors note that s although Big Data is quite successful and popular as a domain, there are still very few cases of its applications on agriculture, especially on small farming, as the numbers of scientific publications and commercial initiatives show. The authors note that the five Vs of Big Data -

The application

Agro 4.0  was presented originally in [8], [28]. Development has continued in this work. In the following session, we introduce the system’s core concepts, briefly describe the system’s architecture as well as the new contributions of this work (hierarchical users grouping, a Simulations module, and visualizations).

Methodology

In this section, we present experimental case studies using the dataset of rural properties collected using Agro 4.0, the system presented in Sections 2 and 4, by a program called Balde Cheio. The data set is made of one register per year by rural property, and it contains the data and fields specified by the ISA Methodology for each property. In Section 5.1, a characterization and analysis of the data set is explored. These analyses are relevant for all further discussion presented in this

Overall dataset analysis and experiments results

In this section, an initial analysis of the dataset is presented and discussed, then the results of both simulations are detailed. The first simulation (Feature Selection, briefly explained in Section 2.5) was done by using the Weka software [32], [33], the second simulation was executed using the Analytical Intelligence tooling developed in this work for the Agro 4.0  software.

In Fig. 9 we can see correlations between the sustainability indicators. Some groups of variables stand out because

Conclusion

Public awareness of the negative impacts of human activity on our environment is at an all times high. Technological efforts to increase the sustainability of productive Agroecosystems are being studied, developed and applied in many different places. In this work, we adopted a Brazilian methodology called Indicators of Sustainability in Agroecosystems (Indicadores de Sustentabilidade em Agroecossistemas – ISA), implemented an information system based on it and apply Data Science techniques

References (67)

  • E.M. De Olde et al.

    Assessing sustainability at farm-level: Lessons learned from a comparison of tools in practice

    Ecol. Indic.

    (2016)
  • N. Van Cauwenbergh et al.

    Safe-a hierarchical framework for assessing the sustainability of agricultural systems

    Agric. Ecosyst. Environ.

    (2007)
  • M. Kropff et al.

    Systems approaches for the design of sustainable agro-ecosystems

    Agric. Syst.

    (2001)
  • R. Shaw et al.

    Characterising the within-field scale spatial variation of nitrogen in a grassland soil to inform the efficient design of in-situ nitrogen sensor networks for precision agriculture

    Agric. Ecosyst. Environ.

    (2016)
  • C.C. de Resende et al.

    Investigating market efficiency through a forecasting model based on differential equations

    Phys. A

    (2017)
  • I. Thysen

    Agriculture in the information society

    J. Agric. Eng. Res.

    (2000)
  • R. Nikkilä et al.

    Software architecture for farm management information systems in precision agriculture

    Comput. Electron. Agric.

    (2010)
  • S. Fountas et al.

    Farm management information systems: Current situation and future perspectives

    Comput. Electron. Agric.

    (2015)
  • B.A. Aubert et al.

    It as enabler of sustainable farming: an empirical analysis of farmers’ adoption decision of precision agriculture technology

    Decis. Support Syst.

    (2012)
  • W.W. Cohen

    Fast effective rule induction

    Machine Learning Proceedings 1995

    (1995)
  • A.M. Novo et al.

    Feasibility and competitiveness of intensive smallholder dairy farming in brazil in comparison with soya and sugarcane: Case study of the balde cheio programme

    Agric. Syst.

    (2013)
  • B.R. Keeble

    The brundtland report: ‘our common future’

    Med. War

    (1988)
  • M. van Marrewijk

    Concepts and definitions of CSR and corporate sustainability: between agency and communion

    J. Bus. Ethics

    (2003)
  • N. Dempsey et al.

    The social dimension of sustainable development: defining urban social sustainability

    Sustain. Dev.

    (2011)
  • J.R. Lamontagne et al.

    Robust abatement pathways to tolerable climate futures require immediate global action

    Nat. Clim. Change

    (2019)
  • D. Tilman et al.

    Agricultural sustainability and intensive production practices

    Nature

    (2002)
  • E. Fonseca et al.

    Agro 4.0: uma ferramenta web para gestão e análise da sustentabilidade em agroecossistemas

    Anais Do XXIII Simpósio Brasileiro de Sistemas Multimídia E Web: Workshops E Pôsteres

    (2017)
  • K. Marzall et al.

    Indicadores de sustentabilidade para agroecossistemas estado da arte, limites e potencialidades de uma nova ferramenta para avaliar o desenvolvimento sustentável

    Cad. Ciênc. Tecnol.

    (2000)
  • E. Commission, Emas - factsheet, 2008, URL...
  • E. Commission, Emas, a premium environmental management toolfor organisations, 2018, URL...
  • E. Commission, Emas and biodiversity, 2016, URL...
  • D. Freebairn et al.

    Reflections on collectively working toward sustainability: indicators for indicators!

    Anim. Prod. Sci.

    (2003)
  • J. Dedrick

    Green is: concepts and issues for information systems research

    Commun. Assoc. Inf. Syst.

    (2010)
  • Cited by (0)

    View full text