Building the Alliance for an Agricultural Data Revolution

The Platform for Big Data in Agriculture is CGIAR’s key initiative to leverage e-research and data-driven impact to reduce poverty, improve food and nutrition security, and support natural resources and ecosystem services worldwide. The Platform, which officially launched in May 2017, aims to effect digital transformations both inside and outside of CGIAR. Scroll down to discover what has been achieved so far.

Building the Alliance for an Agricultural Data Revolution

The Platform for Big Data in Agriculture is CGIAR’s signature initiative to leverage e-research and data-driven impact on the path to reducing poverty, improving food and nutrition security, and improving natural resources and ecosystem services worldwide. The Platform, which was officially launched in May 2017, aims to effect digital transformations both inside and outside of CGIAR. Scroll down to discover what happened during this first year of existence.

YEAR IN REVIEW 2017

The birth of the platform

It is estimated that 2.5 quintillion bytes (2.5 trillion gigabytes) of data are generated each day worldwide. The CGIAR alone surveys 180,000 smallholder farmers every year. However, agriculture lags far behind other sectors in terms of digitalization and all this wealth of information is mostly not used by farmers.

“Better use of data will help drive better policy decisions, helping solve development problems more quickly, cheaply, and at a greater scale than before.”

The Platform was created to democratize decades of agricultural data as a way to empower analysts, statisticians, programmers and others to mine information and develop rapid, accurate and compelling recommendations for farmers, researchers and policymakers.

15 May 2017 The Platform is born

The Platform for Big Data in Agriculture was officially launched at the ICT4D Conference in Hyderabad, India on May 15, 2017 – the culmination of nearly two years of consultation with over 40 private, non-profit, and public stakeholders in digital agriculture worldwide.

As the largest network of agricultural research organizations in the world, CGIAR is uniquely positioned to be a thought leader and global convener on the use of big data and information technology in agriculture. In this way CGIAR will become a broker of big data information, actively promoting data-driven agricultural development.

Building the Alliance

The power of the Platform resides in its broad, global network of partners. Coming from the public and private sectors, and developing or developed countries, experts all over the world share their expertise through the Platform on fields as diverse as agronomy, climate modelling, genomics, analytics, ICT deployment, amongst others.

Platform partners:

The first annual CGIAR Convention on Big Data in Agriculture

On 19–22 September 2017 the Platform convened some 300 global innovators, researchers, and thought leaders from public, private and non-profit partners in Palmira, Colombia for the first annual CGIAR Convention on Big Data in Agriculture.

The Convention opened the way for new collaborations to shape the future of digital agriculture in developing economies.

Explore the network visualization below to find out more our current partners and participants to the 2017 Convention. ↓

Following the Convention, the Platform began discussions with companies offering cloud computing and storage services (exploration of use-cases and collaborations with analytic and cloud services providers including Google, Microsoft, Amazon, and IBM is ongoing), data analytical infrastructures (e.g. University of Minnesota Supercomputing Institute, IBM), machine learning service providers (Google TensorFlow, IBM Watson), and partners providing new pathways to data analysis or impact at scale (e.g. the Digital Impact Alliance to liaise with the mobile industry).

2017 Partnerships Highlights

→ Bridging data gender gap

The Platform has teamed up with CGIAR’s Collaborative Platform for Gender Research to’ co-design innovative uses of data science to help bridge gender divide in data; a key step to ensure women are not left behind in digital agriculture revolution.

→ Private sector partnership for geospatial analysis

The Platform facilitated an alliance with imagery provider Digital Globe, enabling researchers across CGIAR to access high-resolution satellite imagery and run analysis on it in the cloud.

Early Results

This 6-year platform (2017 – 2022) will provide, by implementing initiatives under three Module pillars, global leadership in organizing open data, convening partners to develop innovative ideas, and demonstrating the power of big data analytics through inspiring projects. Under each Module the Platform aims to effect several digital transformations, for both inside and outside of CGIAR, including: mobilizing CGIAR data to accelerate research and spur new data-driven innovations, building collaboration across our organization and with the wider sector, and leveraging CGIAR expertise while claiming a unique leadership voice in digital agriculture.

Module 1: ORGANIZE

The Platform has provided support to CGIAR and partners to fully comply with open-data and open-access principles by addressing technical and organizational challenges. It also addressed critical gaps, both organizational and technical, expanding the horizon of CGIAR research.

Increased data capacity across our 15 partner centers

As a result of allocating a series of grants to each of our 15 partner centers, we have seen a series of inspiring trends emerge as each center implements strategies to mobilise data.

  • Increased investment in developing data repositories and software infrastructure to build open data sharing and storage capabilities
  • Investment in new staff roles for data curation, collection, and analysis, while providing additional training for current staff
  • Reallocation of staff resources to collect, store, and unlock data

GARDIAN

In line with the aim that research data should be findable, accessible, interoperable, and reusable (FAIR), the Platform has built and launched a robust prototype of an open access, searchable data harvester, GARDIAN (Global Agriculture Research Data Innovation and Acceleration Network). The harvester, which spans databases across all CGIAR centers, applies semantic web methods that enables searching across domains. By the end of 2017, GARDIAN showcased about 50,000 publications and 1,800 datasets; it continues to grow, with several new features planned for 2018.

Publications

Datasets

Data Access and Management

The Platform provided increased support and momentum for Centers to comply with CGIAR’s Open Access and Data Management Policy (OADM). Technical guidance and seed funding to all 15 CGIAR Centers in their implementation of this policy contributed to a significant increase (10% or more at most Centers ) in the number of public datasets and publications made available via their respective repositories.

%

increase in public datasets and publications

Responsible data sharing guidelines

While the enthusiasm for data sharing grows, we have been working to ensure that data sharing and use comply with ethical standards that protect those who could be vulnerable to exploitation. We have completed surveys of each of the 15 centers’ privacy and ethics standards, and are in the last stages of developing a set of guidelines to help researchers navigate the evolving implications of technology, confidentiality, intellectual property, consent, access and sharing of benefits.

More community for our Communities of Practice

Our Communities of Practice (CoPs) aim to increase the capacity for problem solving, across the CGIAR and partner network, through collaboration and the sharing of information and knowledge. To increase this working capacity of our CoPs, we established new communications strategies and digital tools that will open up communication channels within and between the communities, with the Platform itself, as well as with the wider scientific community.

Module 2: CONVENE

The platform has brought together big data practitioners with global private sector brands, local entrepreneurs, universities, and others, in spaces that encourage interaction and innovation to solve development problems.

The CGIAR Big Data in Agriculture Convention

The Platform produced the first annual CGIAR Convention for Big Data in Agriculture in September 2017, convening some 300 global innovators, researchers, and thought leaders from public, private and non-profit partners in Palmira, Colombia. The content and attendee list was carefully curated to foster new alliances and advance the discipline of digital agriculture for CGIAR and the sector. The Convention was an important vehicle for recruiting new Community of Practice (CoP) members, and attendees identified critical policy, infrastructure, and investment priorities for the agriculture development sector and contributed to a genuinely multi-stakeholder plan of action for the Platform.

The Communities of Practice (CoPs)

The Platform established Communities of Practice (CoPs) around six relevant topics. For each, key actors across CGIAR and partners to the Platform were identified, and the CoPs started to facilitate improved knowledge management, define collective priorities for driving the science forward and fill critical gaps identified by participants.

Crop Modelling

The Crop Modelling CoP funded three mini proposals to support activities related to modelling that would achieve a demonstrable boost/impact through injecting a small amount of funding. The Crop Modelling CoP started to develop 4 draft review articles that seek to document level-set crop modeling activities and potential impacts from CGIAR centers and partners, with a specific focus on how best to leverage integrated Genotypic, Environmental, Crop Management, and Socio-economic data. During 2017 the CoP grew to 204 members, including CGIAR scientists (40%), research institutes and academia (30%), private sector (20%) and NGOs, government and multilateral organizations (10%).

Visit the COP page →

Socioeconomic

The Socioeconomic CoP formed working groups on gender, ontology-independent structural metadata schema, ethics, privacy and cyber-security, and blockchain. It also identified the need for a socioeconomic data ontology and roadmap for its development. In 2017 the CoP had about 170 members of which a quarter were CGIAR scientists, a quarter academia, a fifth private sector, and the rest NGOs, government, multilateral organizations. Visit the COP page →

Data-Driven Agronomy

The Data-Driven Agronomy CoP identified key actors and initiatives working on data-driven agronomy and gauged strengths and critical gaps across CGIAR and strategic partners, in the process further defined the vision of the CoP. The CoP developed a map of actors working in: weather and climate, soils, crop management and support for extension services. The Data-Driven Agronomy CoP grew to 280 members, a diverse group from the CGIAR centers, research institutes, academic centers, NGOs, public and private sector actors. Visit the COP page →

Geospatial

The Geospatial CoP drafted a position paper on what factors would unlock the potential of precision agriculture in the tropics, assessing its feasibility for crops, livestock, forestry, and fishery. Moreover, the CoP developed the geospatial data cataloging and analysis features of GARDIAN that were soft-launched at the Convention. This tool was developed to serve the functionality of cataloging, visualization, and spatial analysis of large geospatial datasets published by CGIAR. Visit the COP page →

Ontologies

The Ontologies CoP developed new trait ontologies in Crop Ontology and updated the Agronomy Ontology (AgrO) to better integrate with digital fieldbooks. This constituted a critical step for improving the quality and organization of agronomic research data from the point of collection. A collaboration was launched during the Convention with the Socio-Economic Data CoP to start a working group that will address the ontology gap for socio-economic data. The objective supports interoperability of biophysical and socio-economic data, to facilitate such work as studies on the potential adoption of breeding products and looking at gender-and youth-sensitive data and meta-analyses. The Ontologies CoP also engaged several centers in the development of ontologies-based fieldbooks and statistical analysis modules with multiple centers and programs. The CoP grew to 140 members in 2017. Visit the COP page →

Livestock Data

The Livestock Data for Decisions CoP joined at the end of the year. Visit the COP page →

Module 3: INSPIRE

Module 3 consists of putting into practice the opportunities of using open data to solve development problems faster, cheaper and more efficiently.

The Inspire Challenge

In 2017 we challenged partners, universities, and others to use our data to create pilot projects that democratize data-driven insights to inform local, national, regional, and global policies on agriculture and food security in real time. We received more than 120 proposals from applicants from 37 countries.

During our September convention, we awarded grants of US$100,000 to each of the five winning ideas:

Projects wrapped up in-field testing and learning by the fall of 2018, each with encouraging preliminary results. All are seeking pathways and partnerships for replication or scaling

Data-driven innovations from across CGIAR

How’s the water? CIFOR crowdsourced to find out

How’s the water? CIFOR crowdsourced to find out

CIFOR started monitoring water levels and quality in the Sondu Basin of South-West Mau in 2014 using automatic sensors. These generated the first ever precise data set of water flow and water quality information, available continuously over two years. But now, locals have also been trained to read water-level gauges installed along the river and send in their measurements by SMS.

CIMMYT's Lorena Gonzalez fast-forwards action on hunger using technology

CIMMYT’s Lorena Gonzalez fast-forwards action on hunger using technology

Intrigued by the unique relationship our food crops have to their geographical environment, Lorena Gonzalez dedicated her passion for geomatic technology to collect site-specific farm data that is revolutionizing the way researchers and farmers tackle hunger.

ICRISAT awarded Microsoft’s AI for Earth grant to promote environmental sustainability

ICRISAT awarded Microsoft’s AI for Earth grant to promote environmental sustainability

As part of its worldwide AI for Earth program, Microsoft has recently announced grants to provide artificial intelligence (AI) technology to organizations engaged in solving environmental challenges. ICRISAT is one of the grantees and will now be able to use Microsoft Azure resources for weather predictions and pest migration information.

Africa Tree Finder

Africa Tree Finder

This App developed by ICRAF shows the distribution of indigenous tree species and information on the products and services that they can provide. It arms local community members, government agencies, private sector owners, and other land managers with the information they need to select the best tree species for landscape restoration or agroforestry effort. Currently available for Kenya and Uganda.

New ILRI procedure helps to map livestock keeping patterns

New ILRI procedure helps to map livestock keeping patterns

A new mapping procedure developed by scientists at ILRI uses Demographic Health Surveys (DHS) data to bring out spatial patterns of socio-economic factors among livestock herders and systems. The map creation procedure incorporates three innovations: how the DHS data is stored, use of big data approaches to integrate datasets from different countries and use of machine learning.

Free mobile app for rice weed control in Africa

Free mobile app for rice weed control in Africa

AfricaRice launched in April 2018 a powerful new App called ‘RiceAdvice-WeedManager’ to help African rice farmers find the most effective and cost-efficient weed management strategies, matching their specific farming conditions and available resources.

Credits:

Project leader: Marianne McDade
Writing and editing: Marianne McDade and Stefanie Neno
Concept and web implementation: Stefanie Neno