Sander Janssen's presentation during the eROSA Policy Workshop in Brussels, 19 June 2018

1. Community and Governance
Recommendations for the Future
State of an e-Infrastructure in Agri-
Food Sciences
Sander Janssen (WUR), Odile Hologne (INRA),
Panagiotis Zervas (AgroKnow) and many others

2. 2
Link to Roadmap document
Corresponds to
Section 2 on Vision
Section 3 on Grand Challenges
 Section 6 on Recommendations (parts on
community and business models)
Feedback and insights welcome through open
consultation

3. 3
What is/should the community
need?

4. 4
Food System at a turning point
Multiple challenges
Feeding the 9 billion
Climate change
Unhealthy food patterns
Planetary boundaries
Overall challenge = Interconnectedness!

5. 5
Three trends/developments
Adoption of a systems perspective:
More complicated in short term
New genetic techniques
Also/especially for non-commodity crops/breeds
Digital Agriculture (or Data Revolution in
Agriculture)

6. 6
Policy frameworks
• SDGs
• COP21, etc
• Europe2020
• FOOD2030
• EOSC
• Other: IPBES,
etc…

7. 7
Food system in three components
Smart farming, food security & the
environment
Gene-based approaches from omics to
landscape
Food Safety, Nutrition & Health

8. 8
Societal Scientific
• Personalised nutrition and health
advice: advice consumers based on
specific characteristics
• Fast and targeted responses,
preferably ex-ante, to food and health
risks
• Supply chain efficiency across the
actors in the value chain
• Tracking and tracing: transparency
across value chain
• Reducing food waste
• Inclusive and cost-effective health
insurance
• How to connect food intake to health
outcomes? (and to agricultural
production)?
• How to provide estimate & predicts risks
as occurring in the value chain? What
are appropriate responses?
• What are the impacts of changing diets
in terms of food-fuel, protein transition
in relation to the environment, social
conditions and farming?
• What is optimal transparency for a
supply chain? What do consumers
want/need to know?
Food Safety, Nutrition & Health

9. 9
Food Safety, Nutrition & Health
Obstacles Expectations
• Purchasing power in the value chain buys
data access
• Data = power = money
• Lack of mechanisms of benefit sharing across
the supply chain
• Lack of public infrastructures that work along
the supply chain
• Legal validity and governance issues
• Dissemination of scientific outcomes: raising
sensitivity around risks and benefits
• Lack of standardized vocabularies, lack of
standardization.
• Weaknesses in data curation and data rescue
• Better understanding of positive and negative
impacts of openness and sharing
• Urgently need data sharing arrangements
• Need for a broader innovation approach than the
current step in the supply chain
• Demonstrating cases of linked data use and
analytics.
• Collaborative models with the different actors in
the supply chain

10. 10
Example of case study
Food Safety
Impact: Food safety is an integral property of food and
agriproducts that must be secured for health and well
being. The performance of a food supply chain is affected
directly and/ or indirectly by many factors such as climate,
economy and human behaviour.
Beneficiaries: Consumers
Users: EFSA, National Food Safety Authorities, Testing lab’s,
supply chain companies
Role of Science:
 The opportunity to utilise the new technologies being
developed related to monitoring and the use of citizen
science.
 New models will be developed that enables
integration of complex data sets enabling the
prediction of food safety problems at an early stage
Road to open science:
 Models enabling systemic approach of food safety
 Infrastructure (i.e. Web environment) to automatically
collect, process, store, compute, visualise and integrate
data and information from various origins and natures.
 A food safety data platform to collect and connect,
compare and share information about food safety
generated across Europe.

11. 11
Cross cutting issues
Scientific challenge: design methods for
better targeting of farmers/consumers/value
chain actors, while at the same time
improving efficiency, lowering environmental
burdens, improving health
Overall, for the development of Open Science
for food systems, we need to Share, Connect
and Collaborate

12. 12
Share
Across use cases, efforts required in data
curation and data rescue  getting data
available
Beyond data: share analytics, models and the
scientific process
Smarter interoperability platforms: needs to
be easy, not challenging

13. 13
Connect
Be explicit about adopting standards
Use existing ones, do not develop new ones
Recommendations are needed
Establish & advocate ‘best practices’ of open
science
Deliver impact-stories: what does open science
achieve?
Learning resources for capacity building

14. 14
Collaborate
System of systems:
Organize absorption capacity for smaller
projects/initiatives to join
Certify good practices
Innovation incubator: scaling up useful examples
Infra should be as ‘invisible as possible’
Advocate for user centric perspective of EOSC

15. 15
What are our recommendations?

16. 16
General recommendations
Needs to be seen as an ongoing process
Stakeholder inclusive and international
collaborations
Establish sustainable funding mechanisms for
common services
Aligning and integrating with the European
Open Science Cloud and Food2030

17. 17
Specific recommendations
From Open Science to Open Innovation in
the Food System
 Inclusion of public and private players
 Science to innovation: breaking down the silos of
research and private sector
 Open approaches
 Capacity building in open innovation

18. 18
Specific recommendations (2)
Food System e-infrastructures for
international development
Food system challenges are cross
continent/global
 Common global public good
 Involve international organizations (FAO, CGIAR,
CABI, IICA, etc)

19. 19
Specific recommendations (3)
Data-Driven sustainability assessment for
SDG-achievement
 SDGs as a multi-national framework
 Enforce the role of data driven
 Support the role of knowledge systems in data
rich monitoring

20. 20
Specific Recommendations (4)
Open Food System Science for Agri-
Environmental and Nutrition policy making
 Open science as supporting evidence based
policy making
 Joint learning and development across policy-
science interface
Shared infrastructures for experimentation and
policy analysis

21. 21
Specific recommendations (5)
Training the “Food System 4.0” scientists for
open innovation towards Food 2030
 Need for new type of scientists
 working across disciplinary boundaries, data & IT
intensive technologies
 Open interaction, attention for transparency and
reproducibility

22. 22
Specific recommendations (6)
Large Scale Public Private Partnership (PPP)
on data-driven innovation and research
towards Food 2030
Competitive advantage of European Agri-food
data economy
Building links across research, public & private
Powerful analytics, artificial intelligence &
ubiquitous access to data

23. CONSORTIUM
WWW.EROSA.AGINFRA.EU
Thank you for your attention!
@H2020_eROSA

24. 24
Objectives
Identify societal impacts & research challenges that
benefit from an open science e-infrastructure in
agri-food
Identify common challenges in ICT & data that could
be tackled with an e-infrastructure approach
Engage a broad community of scientists with a
diverse background to ensemble transformative use
cases

25. 25
Societal Scientific
• Developing efficient plant and cattle
breeding to provide genetic solutions to
the disruptive changes in food production
• Breeding to support non-intensive farming
(smallholder, organic etc.)
• Speed-up the control of new invasive
species (pests)
• Providing genetic solutions adapted to the
end-user needs (farmers, consumer, etc)
• Helping the development of plant
participatory breedings
• Helping the up-scaling : from omics to
population
• For plant breeding, easy the extrapolation of
results from lab to field(S)
• Improving the characterisation of the
environment components of phenotyping
systems.
• Develop model-assisted breeding
• Providing an alternative to GMOs?
• Opening and sharing data
• Sharing of e-infrastructure (hardware,
software, data repositories etc.)
Gene-based approaches from omics to landscape

26. 26
Gene-based approaches from omics to landscape
Obstacles Expectations
• Available skills to take profit of the open-science
approach
• Shared and adopted international standards
• Starting from problems: having a actual and efficient
user involvement
• Integrate a large diversity (type of data, cultural
differences between omics and higher-scales
communities, IT skills,…
• Having actual interoperable systems
• Involvement of private companies (which business
model, which IP?)
• Available innovation platforms
• Different levels of progress between the plant,
microbiome, and animal communities
• Knowledge gap between current scientific working
practice and Open Science (reg. ICT’s, capacity, IPR,
licensing models etc.)
• Better understanding of positive and negative impacts of
openness and sharing
• Easier to work on broader, cross-domain and cross-
community use cases
• E-infrastructures to not only favour data exchanges and
analysis, but also models and training
• The FAIRification should be transparent
• Better valorisation opportunities (monetizing, citation
etc.)
• Higher virtualisation of the IT system: web services, cloud
=> interoperability, scaling up, traceability, security, etc
• Demonstrating cases of linked data use and analytics.

27. 27
Societal Scientific
• Disruptive changes in food
production without damage to the
less favoured
• Inclusive approach, using local
communities
• Towards new business models –
agriculture as a service
• Support non-intensive farming
(smallholder, organic etc.)
• Fair & sustainable process for farmers
• Balance between supply and
(qualitative) demand, e.g. nutrition
• Responsible ownership of data
• Improving the data value chain
• Using more timely and more
localized data and knowledge
• To be able to serve local
stakeholders and provide more
precise and localized advice
• E-capacity building for intermediaries,
NGO’s, farmers
• Opening and sharing data
• Sharing of e-infrastructure (hardware,
software, data repositories etc.)
Smart farming, food security & the environment

28. 28
Smart farming, food security & the environment
Obstacles Expectations
• Knowledge gap between current scientific
working practice and Open Science (reg. ICT’s,
capacity, IPR, licensing models etc.)
• Lack of incentives to practice OS
• Lack of advocacy and education for Open
Science
• Lack of sharing and re-use culture
• Issue of trust around big data analytics (e.g.
privacy & commercial issues)
• Lack of understanding of business models
• Uncertainty around ownership
• Uncertainty around provenance, traceability,
transparency
• Lack of standards & interoperability
• E-infrastructures to not only support agricultural
production but also the environment, livelihoods
• More respect for and protection of privacy (e.g.
of farmers)
• Grip on data sharing and data protection
• Better valorisation opportunities (monetizing,
citation etc.)
• More collaborative research
• Easier to work on broader, cross-domain and
cross-community use cases
• Better access to better data and data integration
tools
• Improved capacity to work with e-infrastructures
• “reverse science”, using data analytics as the
input for new research

29. 29
Example of case study
Global Agricultural monitoring and early warning systems
Impact: Better predictions of famines, drought and
agricultural production allows for an earlier policy and
disaster relieve response.
Beneficiaries: farmers, rural population
Users: GEOGLAM, policy makers at national and
international level, FAO, UNWFP, development banks,
insurance companies
Role of Science: innovation in the development and
validation of methods and tools required in the fields
of data acquisition, data analytics, modelling and
decision support integrating agronomic, climate, soil
and weather data
Road to open science: Improving the availability of
research infrastructures (HPC, storage, grid),
Improving the availability and access to data and the
capacity to work with Remote Sensing data and other
data sources; Development and testing of big data
analytics solutions for geospatial data.