Forest Variable Estimation and Change Monitoring Solutions Based on Remote Sensing Big Data

In this pilot, we demonstrate the usability of online platforms to provide forest inventory systems for exploiting the benefits of big data. The pilot highlights the technical transferability of online platform based forest inventory services. All of the services tested in the piloting sites were technically implemented successfully. However, in new geographical areas, strong user involvement in service definition and field data provision will be needed to provide reliable and meaningful results for the users. Overall, the pilot demonstrated well the benefits of technology use in forest monitoring through a range of forest inventory applications utilizing online big data processing approaches and inter-platform connections.

Future Vision, Summary and Outlook

The DataBio’s agriculture pilots were carried out through a multi-actor whole-farm management approach using information technology, satellite positioning and remote sensing data as well as Internet of Things technology. The goal was to optimize the returns on inputs while reducing environmental impacts and streamlining the CAP monitoring. Novel knowledge was delivered for a more sustainable agriculture in line with the FAO call to achieve global food security and eliminate malnutrition for the more than nine billion people by 2050. The findings from the pilots shed light on the potential of digital agriculture to solve Europe’s concern of the declining workforce in the farming industry as the implemented technologies would help run farms with less workforce and manual labor. The pilot applications of big data technologies included autonomous machinery, mapping of yield, variable rate of applying agricultural inputs, input optimization, crop performance and in-season yields prediction as well as the genomic prediction and selection method allowing to cut cost and duration of cultivar development. The pilots showed their potential to transform agriculture, and the improved predictive analytics is expected to play a fundamental role in the production environment. As AI models are retrained with more data, the decision support systems become more accurate and serve the farmer better, leading to faster adoption. Adoption is further stimulated by cooperation between farmers to share investment costs and technological platforms allowing farmers to benchmark among themselves and across cropping season.

Conclusions and Outlook—Summary of Big Data in Forestry

In this chapter, we summarize the findings from the forestry pilots conducted during the DataBio project. Although the pilots demonstrated the functionality of big data in forestry through several practical applications and services, they also highlighted areas where further development is needed. More effort is needed particularly in ensuring smooth connections between the technical components of the processing pipelines, as well as designing the best business solutions within the big data service chain and between the service providers and users. Overall, the challenge for the coming years is to establish operational big data processing pipelines that meet the requirements and expectations of forestry stakeholders.

Crowdsourced Data

Crowdsourcing together with Volunteered Geographic Information (VGI) are currently part of a broader concept – Citizens Science. The methods provide information on existing geospatial data or is a part of data collection from geolocated devices. They enable opening parts of scientific work to the general public. DataBio Crowdsourcing Solution is a combination of the SensLog server platform and HSLayers web and mobile applications. SensLog is a server system for managing sensor data, volunteered geographic information and other geospatial data. Web and mobile applications are used to collect and visualize SensLog data. SensLog data model builds on the Observations & Measurements conceptual model from ISO 19156 and includes additional sections, e.g., for user authentication or volunteered geographic information (VGI) collection. It uses PostgreSQL database with PostGIS for data storage and several API endpoints.

Linked Data and Metadata

Data is the basis for creating information and knowledge. Having data in a structured and machine-readable format facilitates the processing and analysis of the data. Moreover, metadata—data about the data, can help discovering data based on features as, e.g., by whom they were created, when, or for which purpose. These associated features make the data more interpretable and assist in turning it into useful information. This chapter briefly introduces the concepts of metadata and Linked Data—highly structured and interlinked data, their standards and their usages, with some elaboration on the role of Linked Data in bioeconomy.

Data Pipelines: Modeling and Evaluation of Models

This chapter outlines the utility of data pipelines modeling in the context of a data driven project and enumerates metrics for evaluating the quality of the data modeling regarding the readability and the comprehensibility of the models. We start with explaining the challenges surrounding the DataBio project that led to the adoption of data pipelines modeling using the Enterprise Architecture language ArchiMate. Then we present the data modeling process with examples from DataBio pilot studies starting with modeling software components provided by project stakeholders and ending up with integration of components into data pipelines that achieve the data analytics lifecycle intended by the pilot study. We end the chapter with the evaluation of the quality of DataBio data pipelines models with metrics collected by a monitoring tool for ArchiMate models.

Real-Time Data Processing

To remain competitive, organizations are increasingly taking advantage of the high volumes of data produced in real time for actionable insights and operational decision-making. In this chapter, we present basic concepts in real-time analytics, their importance in today’s organizations, and their applicability to the bioeconomy domains investigated in the DataBio project. We begin by introducing key terminology for event processing, and motivation for the growing use of event processing systems, followed by a market analysis synopsis. Thereafter, we provide a high-level overview of event processing system architectures, with its main characteristics and components, followed by a survey of some of the most prominent commercial and open source tools. We then describe how we applied this technology in two of the DataBio project domains: agriculture and fishery. The devised generic pipeline for IoT data real-time processing and decision-making was successfully applied to three pilots in the project from the agriculture and fishery domains. This event processing pipeline can be generalized to any use case in which data is collected from IoT sensors and analyzed in real-time to provide real-time alerts for operational decision-making.

Big Data Visualisation

In this chapter, we introduce the topic of big data visualization with a focus on the challenges related to geospatial data. We present several efficient techniques to address these challenges. We then provide examples from the DataBio project of visualisation solutions. These examples show that there are many technologies and software components available for big data visualisation, but they also point to limitations and the need for further research and development.

Copernicus Data and CAP Subsidies Control

This chapter integrates the results of three pilots developed within the framework of the Horizon 2020 DataBio project. It aims to provide a broad picture of how products based on Earth Observation techniques can support the European Union’s Common Agricultural Policy requirements, whose fulfillments are supervised by National and Local Paying Agencies operating in Romania, Italy and Greece. The concept involves the use of the same data sources, mainly multitemporal series of Copernicus Sentinel-2 imagery, but through three different Big Data processing chains, tailored to each paying agency’s needs in terms of farm compliance assessment. Particularities of each workflow are presented together with examples of the results and their accuracy, calculated by validation against independent sources. Business value aspects for each use case are also discussed, emphasizing the way in which the automation of the CAP requests verification process through satellite technologies has increased the efficiency and reduced cost and time resources for the subsidy process. We end the chapter by highlighting the benefits of continuous satellite tracking as a substitute, but also complementary to the classical field control methods, and also the enormous potential of Earth Observation-based products for the agri-food market.

Introduction—State of the Art of Technology and Market Potential for Big Data in Forestry

Forest monitoring is undergoing rapid changes due to the growing data volumes, developing data processing technologies and increasing monitoring requirements. The DataBio forestry pilots set out to demonstrate how big data approaches can support the forestry sector to get full benefit of the evolving technologies and to meet the increasing monitoring requirements. In this introductory chapter, we describe underlying technical and market forces driving the forestry sector toward big data approaches, and give short overviews on the forestry pilots to be presented in the following chapters.