Invited talk, PHIL_OS, March 30-31 2023, Exeter
https://opensciencestudies.eu/whither-open-science. Includes hidden slides.
FAIR and Open Science needs Digital Research Infrastructure, which is a federated system of systems and needs funding models that are fit for purpose
Culture change needed for paying for Open Science’s infrastructure and funding support for data driven research needs more reality and less rhetoric
Can’t Pay, Won’t Pay, Don’t Pay: Delivering open science, a Digital Research Infrastructure perspective
1. Can’t Pay, Won’t Pay, Don’t Pay
Delivering open science,
a Digital Research Infrastructure perspective
Professor Carole Goble CBE FREng FBCS
Software Sustainability Institute UK
ELIXIR, ELIXIR-UK Head of Node
IBISBA, Head of Digital Infrastructure
The University of Manchester, UK
carole.goble@manchester.ac.uk
PHIL_OS, March 30-31 2023, Exeter
https://opensciencestudies.eu/whither-open-science/
2. https://en.unesco.org/science-sustainable-future/open-science/recommendation *UKRI: The UK’s research and innovation infrastructure: opportunities to grow our capability, 2020
FAIR and Open Science needs
Digital Research Infrastructure
• Scientific equipment: sensors, instruments;
• Knowledge and data services: collections, archives and
scientific data; catalogues & registries; collaboration tools;
electronic lab notebooks;
• Software: specialist tools, libraries, workflows, platforms;
• Core services: data and computing systems, storage,
communication networks;
• Commons services: AAI, PID management, search;
• People: DevOps, Research Software Engineers, Data
Stewards, support desk, trainers, project managers.
3. ELIXIR: a digital research infrastructure
for Life Science Data
https://elixir-europe.org
23 nodes
220+ orgs
Hub
Data, analysis and
people mobilisation
National COVID-19 Data Portals
4.
5. ELIXIR Digital Infrastructure Services
Trusted curated discipline
data archives
National RDM
platforms & support
FAIR Commons Services
& Registries
Data analytics
&Tools
Professionalisation of
Software & Data
Knowledge Management
andTraining
47 21
282
27 20+
128
30+
6. Generalist Repositories
Public (cloud)
Infrastructures
Commercial services
Global, public discipline archives
Community/discipline
infrastructure
National
Infrastructure
Digital Research Infrastructure Ecosystem
Different
scales
Different ownership
Different
business
models
Independent but
Inter-dependent
Ideally friction free
with quality content
Institutional
Infrastructure
Personal/project infrastructure
(server under the desk, project web site, group gitlab)
7. Open Science & DRI “System of Systems”.
To Build. To Operate. To Sustain.
Image: https://element.io/features/closed-federation-and-open-federation DOI: 10.5281/zenodo.1064730
Infrastructure ecosystems and federations need time, cooperation and glue
8. “Data sharing is not a simple matter of individual practice, but one of
infrastructure, institutions, and economics.”
Borgman, C. L., & Bourne, P. (2022). Why It Takes a Village to Manage and Share Data.
Harvard Data Science Review, 4(3). https://doi.org/10.1162/99608f92.42eec111
https://www.science.org/doi/epdf/10.1126/science.add2734
Science 22 Dec 2022 Vol 378, Issue 6626
9. • Lack of funding
• Resistance to change
• Data privacy & security
• Technical barriers
• Interoperability
• Legal and ethical
• Skills and training
“requires a concerted effort
from the research community,
including funding agencies,
institutions, and researchers, to
work together”
11. “All agencies will fully
implement updated policies,
including ending the optional
12-month embargo, no later
than December 31, 2025.”
https://www.nature.com/articles/d41586-022-00402-1
Nature, Feb 2022
https://www.science.org/content/article/ready-set-
share-researchers-brace-new-data-sharing-rules
Science 379(6630) Jan 2023
13. What are the costs, and who will pay?
Science 379(6630) Jan 2023
University burden
• “another “unfunded mandate” from the federal government”
• ”special burden for smaller institutions”
Grant change management – PI and Funder
• “NIH has a strict dollar limit for many grants, data-sharing costs
may cut into the funds available for research.”
Digital Infrastructure
• “Even the largest repositories are still looking for sustainable
business models.”
• “Discipline-specific ones are typically supported by grants for
individual projects that don’t assure funding after the grant ends.”
They referred to data
curation costs … isn’t
that a necessity not a
luxury?
14. DRIs span funding, project and organisational boundaries
Core funding fade Multiple independent
funding streams
Collaboration considered necessary
Spread the burden
Infrastructure delivery
coordination*
*Figure Credit: Christopher Woods, Solving the long-term maintenance and
funding challenge of research software by founding the OpenBioSim
Community Interest Company, RSECon2022,
15. How do DRIs get funding?
Donut funding
Add features,
hide maintenance
Core Funding
Infrastructure grants
Sponsorship
Borrowed funding
Research project,
top slice development
Fees, Subscriptions, Donations
Hard to handle cash flows,
especially in universities
Diversification
Spread the risk
In kind labour
Volunteers
Hand shake agreements
Donations by institutions
Commercial platform subsidy
Direct
resourcing
Indirect
subsidies
16. Figured out a business model yet?
https://doi.org/10.1787/302b12bb-en.
18. No mechanism to pay, e.g. Capital funding only so no cloud computing or people
Funding silos for pan-project, pan-organisation infrastructure e.g. X has the Skills but Y
has the money
Culture of free access to data and software, though do pay for access to
instruments, facilities and compute.
PIs unrealistic or no costings into grant awards.
Hidden cost of curation - PhDs or data steward professionals
Can’t Pay?
Don’t Pay?
19. “We only fund research”. Finding funding sources for necessary “boring” stuff and
sustaining infrastructure and getting it used rather than new and shiny thing.
“We will fund a new shiny AI tool but we won’t fund it to interoperate with
anything else, or fund the data to be AI-ready”
“everyone realises that Research Data Management is important, but no-one wants
to pay for it” The evolving landscape of Federated Research Data Infrastructures, 2017
Won’t Pay?
20. https://xkcd.com/2347/
User facing shiny thing
Applications, tools, scripts
Domain specific reusability
Underware
Platforms, infrastructure, libraries
Infra glue – like standards & APIs
Cross-domain generic reusability
Overly familiar
The Pay Gap
21. A UK AI Research Resource
An Exascale capability
https://www.gov.uk/government/publications/future-of-compute-review/the-future-of-compute-report-of-the-review-of-independent-panel-of-experts
22. Commercial players and subsidies
No such thing as a free resource…
Build
Borrow
Buy
https://blog.alexellis.io/docker-is-deleting-open-source-images/
Total Cost
23. “Science has been too singularly focused on innovation, otherwise we would not see
the maintenance of foundational digital infrastructure as a lesser task within the
hierarchy of academic outputs….
…..each and every funder needs to speak up and act, or we risk our silence being
interpreted by a generation of data scientists and software engineers as an
endorsement of the status quo.”
25. Sponsorship – non-project
based organisations & societies
National and pan-national initiatives
Institutional organisations
Societies
Organisations encouraged to celebrate
open source contribution
Collective responsibility
Takes a Village
Borgman, C. L., & Bourne, P. (2022). Why It Takes a Village to Manage and
Share Data. Harvard Data Science Review, 4(3).
https://doi.org/10.1162/99608f92.42eec111
26. What can we do to
normalise and
incentivise paying
for the DRIs that
enable
open science?
Emma Henderson, UKRN workshop on Indicators for Open Research, 15 March 2023
27. Trust <-> Sustainability Reciprocity and Values
Credit for infrastructure providers
not just credit/citation for content providers….
Users must trust providers – to make
infrastructure available, usable,
reusable & fit for purpose.
Providers must trust users - to
promote, advocate, acknowledge and
help sustain it. Support it or lose it.
“the acknowledgement ‘count’ as a Key
Performance Indicator will then be ensured of
an accurate basis for attributing funds, which
in turn impacts a core facility’s ability to invest
in equipment and salaries.”
Otherwise, its text mining mentions!!!
28. Open Research Infrastructure is
Not Free & it is to do with Me
Free labour is Slavery
If there isn’t money, there must be fame. You cannot rely on love.
“Effectively, the community at large are free-riding on the time and
resources invested by open-source developers to maintain critical digital
tools.” Knowles, Mateen, Yehudi, Nature Computational Science 2021
Image: Fig. 119 Making your first pull request on GitHub. The Turing Way project illustration by Scriberia. Used under a CC-BY 4.0 licence. DOI:
10.5281/zenodo.3332807.
29. Get researchers and their communities to care about service providers.
Long-term availability of research data/software/infra, and its development,
to properly feature in research evaluation and valued
Open Research Infrastructure is
Not Free & it is to do with Me
Hans Pfeiffenberger@RDA2o
30. Funded mandates …
5% top slice across all grant awards to support DRI
Barend Mons, Ex President CODATA, President GO-FAIR
% of a research grant award ringfenced to support data & DRI
Andrew Trelor, Australian Research Data Commons
% APC charges go to DRI
Kathleen Shearer, Confederation of Open Access Repositories
31. DRIs can make it easier to love them,
to contribute to them & be aware of them
• “Customer delight”
• Researcher-centric productivity
• Seamless integration (yeah, but who
funds that?)
• Google, Amazon, Microsoft good at this
(and not hampered by funding-driven
architecture designs)
• Support mechanisms
• Social Enterprises: Not for Profits,
Foundations, Community Interest
Company, Fiscal Sponsors
32. Policy:
make it required
Incentives:
make it rewarding
Communities:
make it normative
Skills/training:
make it easy
Infrastructure:
make it possible
Adapted by Michelle Barker from Brian Nosek, Strategy for Culture Change (2019)
Culture Change for Open Science
33. Policy:
make it required
Incentives:
make it rewarding
Communities:
make it normative
Skills/training:
make it easy
Infrastructure:
make it possible
Adapted by Michelle Barker from Brian Nosek, Strategy for Culture Change (2019)
Culture Change in Paying for Open Science’s Infrastructure