European Perspectives on Open Science Policy/JC Burgelman

SA-EU OPEN SCIENCE WORKSHOP PROGRAMME

Johannesburg 30 November – 1 December 2017

JC Burgelman,

DG RTD

European Perspectives on Open Science Policy

1. Why Open Science Policy

2. The 8 Open Science policy priorities

3. The Open Science policies so far

3.1 Open Access to publications

3.2 Open Access to Data

3.3.European Open Science Cloud

3.4 Rewards and Skills for Open Science

3.5 Research Integrity in Open Science

3.6 Open Science and citizens science

4. Mainstreaming Open Science: H2020 & FP 9

5. Conclusion: no way back but much more needed

Open Science = Systemic transition of science system which affects the way

• research is performed• knowledge is shared/diffused/preserved• research projects/results are evaluated• research is funded• researchers are rewarded• future researchers are trained

Affecting the whole research cycle and all its stakeholders

A typical techno-economic paradigm shift a la Perez (technology, market and institutional change go hand in hand)

or to put it different: disruptive and hence disturbing….

1. Well accepted

A general model for open science in early stage drug discovery

Public-Private

PartnershipPublic Domain Commercial

Tools & Basic KnowledgeNOVEL Proteins only!

• Structure

• Chemistry• Antibodies• Screening• Cell Assays

Discovery and Exploration

• No patent• No restriction on use• Open access to tools and data.• Target identification & validation

Drug Discovery and Development

Facilitated by access to increased amount of information in the public domain

- (re)Screening- Lead Optimisation- Pharmacology- Metabolism- Pharmacokinetics- Toxicology- Chemical development- Clinical development

CREATIVE COMMONS PROPRIETARY

Weigelt J. EMBO Reports 10:941-5 (2009)

1. Because we see the benefits

Structural gnomics consortium

1. Benefits for ALL sciences

1. And we need more

Rise of interdisciplinary, highly collaborative international big science projects – open by design

1. 90% OF ALL SCIENTIFIC ARTICLES (2 MILLION PER YEAR) ARE BEHIND PAYWALLS! (AND 1 MILLION PAPERS ARE NEVER READ);

UNIVERSITIES WORLDWIDE WOULD SAVE 6 BILLION DOLLARS IF ALL PAPERS WOULD BE PUBLISHED IN OA (MAX PLANCK REPORT)

2. ONE MILLION VALID SCIENCE PAPERS ARE BOUNCED BACK (REJECTING, RESUBMISSION ETC)

TO DEAL WITH THAT: 100 MILLION HOURS (OF SCIENTIST!) ARE LOST ANNUALLY IN THE ACTUAL PROCESS OF PAPER REJECTING, RESUBMITTING, WAITING ETC.;

SO: MORE OPEN ACCESS, LESS LOSS

3. BIG PRIVATE RESEARCH TRUSTS GO FOR RADICAL OPEN ACCESS: WELLCOME TRUST, GATES, ZUCKERBERG , ETC

4. EVERY 3 DOWNLOADS OF A FRONTIERS ARTICLE LEADS TO 1 CITATION:

IN OTHER WORDS: NUMBER OF DOWNLOADS IS GOOD FOR YOUR IMPACT FACTOR IN OTHER WORDS: OPEN ACCESS IS GOOD FOR YOUR CAREER

1. Plus - it pays off

• Better ROI of the R&I investments: self evident: if

all the results of our public research are made reusable, it will follow that better use is made

• Faster circulation of new ideas: we have 22 million EU

SME's that will have access to top notch research without having to significantly pay for it!

• More transparency of the science system as such: the public taxpayer has this right

• Fit for 21st century science purpose: all grand societal

challenges NEED cross disciplinary research

For researchers:

• Wider dissemination and sharing of the results

• More visibility and credit for those collecting and sharing underlying research data

• New career paths e.g. data scientists, start-ups, science diplomacy

1. Open Science offers great challenges & opportunities 4 science, scientists & society

1. The genesis and context of Open Science Policy











" As I see it, European success now lies in sharing as soon as possible, (…). The days of open science have arrived."

Speech at "Presidency ConferenceOpen Science", 04 of April, 2016,Amsterdam

2. Top level policy

2. Top level policy at national level too

• The Netherlands, first country to make the goal of default OS part of a new (Rutte III) government declaration

"Open science and open access will become the norm in academic research"

• Finland: created a cross cutting OS policy approach

• Several Lander having dedicated platforms or approaches

• VRWI in Flanders comes up with consolidated advice

• G7

• RSA!

• Etc..

2016 - Holistic Policy Agenda: scope & ambitions

… 4 with regard to the use & management of research results and data

Open Data: FAIR data sharing is the default for funding scientific research

Science cloud: All EU researchers are able to deposit, access and analyse European scientific data through the open science cloud, without leaving their desk

Altmetrics: Alternative metrics to complement conventional indicators for research quality and impact (e.g. Journal Impact Factors and citations)

OA & Future of scholarly communication: All peer reviewed scientific publications are freely accessible

12

2. EC Policy Priorities

… 4 with regard to relations with research actors (researchers, institutions and funders)

Rewards: The European research career evaluation system fully acknowledges Open Science activities

Research Integrity: All publicly funded research in the EU adheres to commonly agreed Open Science Standards of Research Integrity

Education and skills: All young scientists in Europe have the necessary skills and support to apply Open Science research routines and practices

Citizen Science: CS significantly contribute and are recognised as valid knowledge producers of European science

13

2. EC Policy Priorities












Then and now

FP7

OA Green orGold+GreenPilot

H2020

OA Green orGold+Greenobligation

& ORD Pilot

H2020

OA Green orGold+Greenobligation

& ORD by default

3. OA: Considerable progress in 10 years made

3.1 OA: key issues Now

• Open Access to Publications: since 2014 mandatory

Set up of an Open Access Publications Platform: stand alone peer reviewed scientific articles resulting from H2020 projects (2017-2018)

• Open Access to research data: default from 2017 onwards. - Opt-outs possible at any stage (IPR, personal data protection and national security)- Research Data Management Plans: mandatory and FAIR

• Mainstreaming in all MS

• Set up of the European Open Science Cloud: a trusted virtual environment for enabling data driven science across boundaries and disciplines (2017-2019)












• May 2016 Competitiveness Council Conclusions: fullopen access to scientific publications by 2020 –commitment of the EU for all Member States!

• Amsterdam Call for Action on Open Scienceadvocates for "full open access for all scientificpublications",

• The Commission wants to lead by example inHorizon 2020:

• - Currently 60-68% of scientific publications produced are open access

further measures needed to achieve target

3.1 OA to publications: political support

3.1 “Sharper” positioning

3.1 Even for the “top”

Towards a EC Horizon 2020 Open Research Europe Publishing Platform

Our proposal

(current state of thinking)

3.1 We can’t debate forever

An online platform allowing rapid, OA publication of

(i) H 2020 related peer reviewed articles; and

(ii) H 2020 related pre-prints which meet basic criteria onauthorship, non-plagiarism and ethical conduct

• contains mechanisms for open/collaborate/public peerreview and a suit of innovative ('next generation') metrics

• is not a repository, it provides a fast, cost efficient andhigh quality service to publish in the 21st century

• is intended for Horizon 2020 beneficiaries as a free,complementary service and is thus not compulsory.

3.1 What is ORE

• March 2017 – CssR Moedas informally discussed the idea with the Open Science Policy Platform (stakeholders)

• May 2017 - CssR Moedas introduced the idea in the Compet Council

• Ongoing - Implementation of ORE through a call for a public procurement in the WP 2018:

Call for tender open in December 2017

Applications until early 2018

Contract signature during 2018

3.1 Timeline












3.2 Now: Open Data

• Open Access to research data: default from 2017 onwards. - Opt-outs possible at any stage (IPR, personal data protection and national security)- Research Data Management Plans: mandatory and FAIR

• Mainstreaming in all MS

• Set up of the European Open Science Cloud: a trusted virtual environment for enabling data driven science across boundaries and disciplines (2017-2019)

3.2 Open Data: basic principle

1. The genesis and context of Open Science Policy2. The 8 Open Science policy priorities3. The Open Science policies so far3.1 Open Access to publications3.2 Open Access to Data3.3.European Open Science Cloud3.4 Rewards and Skills for Open Science3.5 Research Integrity in Open Science3.6 Open Science and citizens science4. Mainstreaming Open Science: H2020 & FP 95. Conclusion: no way back but much more needed

European Open Data (and data drivenscience) supported by the European

Open Science Cloud

3.3 EOSC

29

Source: RTD

The EOSC will allow for universal access to open research data and create a new level playing field for EU researchers

CERN, EMBL,

ELIXIR, etc.

Institutional

repository

Member State

Infrastructure

New provider/

service

Researcher • Easy access through a universal

access point for ALL European

researchers

• Cross-disciplinary access to data

unleashes potential of

interdisciplinary research

• Services and data are

interoperable (FAIR data)

• Data funded with public money is

in principle open (as open as

possible, as closed as necessary)

1.Access to all European research data

2.Access to world-class data services

3.Clear rules of use and service provision

4.FAIR data tools, training and standards

5.Sustainable after the grant

Seamless environment and enabling interdisciplinary research

29

3.3 EOSC and the researcher

30

• Vision is now as clear to external stakeholders

• Single online platform where all European researchers will be able to:

• Find, access and re-use data produced by other scientists.

• Deposit, analyse and share data they have been paid to produce.

• Initially (until 2020), the EOSC will build on existing investments, no/little fresh money is needed

• The Commission provides top-up money to set up and organise the federation and to start creating common European resources.

• EOSC will provide 1.7m EU researchers an environment with free, open services for data storage, management, analysis and re-use across disciplines.

• EOSC will JOIN existing and emerging horizontal and thematic data infrastructures, bridging todays fragmentation and ad-hoc solutions.

• EOSC will add value (scale, data-driven science, inter-disciplinarity, faster innovation) and leverage past infrastructure investment (10b per year by MS, two decades EU investment).

From Vision …. … to Action

After the EOSC Vision has been adopted and embraced by the relevant stakeholders, it is now time for action

Source: RTD

Vision endorsed by the EP, by the EESC and the CoR, by the G7 and “copied” by a host of nations globally: Japan, Canada and China.

30

3.3 Open Data and EOSC

o 110 key participants

o 80 from all scientific fields

o 15 national scientific infrastructures

o 13 research funders

o 19 officials from Member States and

Associated Countries

o Overall, 23 Member States and Associated

Countries represented

o 1800 via web stream and extensive coverage

via Twitter

3.3 EOSC summit

So far:

Stakeholder endorsements: 69

Stakeholder commitments: 61

• The Summit provided strong support for the implementation of the EOSC and marked a step change in the initiative

• Ground European science in a common culture of data stewardship & sharing throughout research data lifecycle. Only a considerable cultural change will enable long-term reuse of research data

• Develop the EOSC as a commons of research data, knowledge and infrastructure with different roles and responsibilities by actors at EU and MS level

• Stimulate compliance through incentives and rewards

3.3 EOSC Summit Highlights 1

● Make FAIR principles pragmatic, equally encompassing all four dimensions: findability, accessibility, interoperability and reusability

● Apply FAIR principles to all digital research objects, incl. data-related algorithms, tools, workflows, protocols & services

● Disciplines must develop their notion of FAIR in a coordinated fashion. Standards are fundamental but a one-size-fits-all approach must be avoided


● Build trust between all stakeholders, e.g. scientific communities, e-infrastructures, research infrastructures, funders - “look outside the organisational boxes and work together”

● Set out both the science case and the financial case to raise commitments for the EOSC, in particular of MS – “whatever we do needs to be integrated with the national systems”

● Some of the actions identified will require more times than others to implement due to e.g. budget commitments, division of labour, building of trust. Implementation will need to fast-track some actions over others, depending on the level of priority, support and maturity

● Develop a formal framework for governing the EOSC (open, dynamic, trial-and-error process) to sustain and strengthen related policies & programmes and secure commitment of funders and users -“what needs to be governed and how?”


• Hundreds of stakeholders involved (unlike when the

internet was created)

• Billions of investment touched: overall annual investment

of 10 billion euros in the EU (mainly by MS) on Research

Infrastructures and e-Infras

• And much more to come: If all public funders would agree

to commit 1% of the overall spending to open

data…(approx 2 billion per year)

3.3 EOSC implementation process: complex but












Working Group on ‘Skills and competences’

Objective

• Ensure that researchers have the appropriate skills to practice Open Science

• Open Science skills shall become an integral and streamlined component of education, training and career development

Report: published in July 2017, discussed at OSPP on 13.10.17

https://ec.europa.eu/research/openscience/index.cfm?pg=home

EU Skills & qualification matrix for OS could be used for this purpose, taking intoconsideration:

• what are the necessary training and competences requirements for aresearcher

• The relevant supporting structures to engage researchers in Open Science

3.4 Rewards & skills


Working Group on 'Evaluation of research careers fully

acknowledging Open Science practices'

Objective :

• Modernisation of the current researchers' career evaluation system

• Creation of incentives and rewards for researchers engaged in Open Science

Report: published in July 2017, discussed at OSPP on 13.10.17


An OS Career Assessment Matrix (OS-CAM) : could be used for this purpose, taking into consideration - cf. abstract report :

38

3.4 Rewards and skills













Council Conclusions on Research integrity (01/12/2015)

The Conclusions stress the relationship between research integrity and OpenScience.

"Recognises the importance of open science as a mechanism for reinforcing research integrity, while, at the same time, research integrity contributes to open science"

Responding to the Conclusions the Commission :

• Initiated the revision of the European Code of Conduct of Research Integrity.ALLEA adopted the new Code on 24 March 2017.The new code supports open science by promoting open access to researchpublication and data as well as the FAIR principles

• The Horizon 2020 legal basis will be adapted (article 34 of the Model GrantAgreement) to be aligned with the revised Code (end 2017).

3.5 Research Integrity












• An EC citizen science strategy being developed: aiming at supporting citizen science in FP9; working towards a citizen-friendly R&I governance framework; and ensuring optimum coordination within EC services.

• A (substantial) EC funding made available:

SwafS-15-2018-2019: Exploring and supporting citizen science (CSA/RIAs, EUR 16m)SwafS-17-2019: Consolidating and expanding the knowledge base on citizen science (RIA, EUR 2.5m)

3.6 Citizen science












4. As a funder: OS in H2020

4. As a funder: OS in FP9beyond OA, ''do'' OS

FP9 goes beyond OA (publications and data) to embrace & incentivise Open Science as real life modus operandi

• FP9 clarifies and strengthens the OA obligations

• FP9 empowers authors of scientific publications

• FP9 is home of FAIR data sharing while complying with IPR rules and exploitation obligations set in the GA

• FP9 broadens OA (with opting out options) to other research output

• FP9 promotes compliance with 'Open Science principles' through a combination of obligations and incentives

• FP9 further implements sanctions for those beneficiaries thatrepeatedly and consistently fail to provide the required open access, requiring institutions to assume responsibility for their intellectualoutput

• FP9 introduces the use of new innovative ('new generation') metricsfor better assessing the impact of research output and the engagement in Open Science

Note: where the case, individual opt-out options remain.













The EC is delivering what it promised:

Substantial progress on the 8 priorities (next generation metrics foreseen for 2018 – advice OSPP ready)

Rigorously in co-design and co-creation with the key stakeholders

ExternallyOS Policy Platform - Mobilising the most important external stakeholders

InternallyOS Policy Task Force – mainstreaming inside the house

OS is here to stay:

It has often been said that 'data is the new oil'. This has been helpful in stressing the value of data in our economy, but it is the wrong analogy. Contrary to oil, open research data is non-exclusive and non-rival and so it makes sense to have publicly funded research data openly available. Open data is more like a renewable energy source: it can be reused without diminishing its original value, and reuse creates new value. The EC’s policies on open research data simply guarantee as many scientists and innovators as possible make, under equal conditions, the best use of this renewable source.

Digital Science Report.The State of Open Data 2017

Stefaan G. Verhulst and Andrew Young (2017)

Open Data in Developing Economies. Cape town: African Minds

1.

5. Much more needs to be done

OS is here to stay:

1.If you want to go fast, go alone.If you want to go far, go together

(African saying)

But also

Such a long journey ahead of us

(South African writer, A. Brink)

5. Much more needs to be done

5. Indeed – the journey only started

User-centered Publishing delivers Precision Information

The Machine is the New Reader

Science as a Social Machine

Data Privacy requires a Web of Trust

Big Data meets Artificial Intelligence

50

• Due to the power of cyber science tools, it is quite realistic to assume that we will evolve from peer reviewed open access publications

to peer reviewed open access research workflows

• Implying that scientific publishers become open science platforms in which an article is 1 of the many products (and not even per se)

• And where the whole process is open and peer reviewed(see the Jacobson controversy --- NAS)

5. e.g. OA of articles the final frontier?

5. No one foresaw this exponential growth either

5. Independent QUALITY assurance will always beat the core of science

More information at

http://ec.europa.eu/research/openscience

55

http://ec.europa.eu/research/openscience

Open Science Monitor

http://ec.europa.eu/research/openscience/monitor/

http://ec.europa.eu/research/openscience/monitor/

Open science represents an approach to research that is collaborative, transparent and accessible. Open science occurs acrossthe research process and there are many different activities that can be considered part of this evolution in science. The open science monitor tracks trends in areas that have consistent and reliable data.

Open science monitor

Explore open science characteristics and indicators.

Home About Open access Open research dataOpen scholarly communication

Citizen scienceDrivers and

barriers

57

* These indicators are for both open access and open scholarly communication.

58