How can repository providers ensure that the data submissions are FAIR? How could the technical support look like?

General consideration: The meaning of Findable and Accessible for (meta)data is technically clear, while Interoperable and Re-usable are unclear.

1. Define what FAIRness means on a technical basis?
   • Consider the level of FAIRness from a scientific perspective (to prevent fulltime documentation).
   • Implement community standards.
   • Design a seamless technical connectivity and develop workflows that allow automatic workflows.
   • Connect curation services to services and repositories.
2. Communicate clear requirements for technical solutions
   • Invest resources in community management.
   • Provide guidance for researchers (e.g. support, training) (>> GO TRAIN)
   • Offer a compliance check during the publishing process.
   • Establish metrics and level of FAIRness.
   • Ensure readability by humans.
3. Change
   • Create technology that rewards FAIRness and documentation efforts immediately, e.g. citation, credits. (>> GO CHANGE)
   • Establish a workflow reaching from managed data to long-term availability.
   • Make use of feedback to improve and develop standards continuously.

Which technical framework or support services are required to generate FAIR data? (from a researcher’s perspective)

General consideration: Offer recommendations for action by drawing a road to FAIR and by explaining general objectives and related processes.

• Develop and apply consistent policies and RDM processes at every institution.
• Establish interfaces between institutions.
• Provide digital workflows and systematic IT support to close the competency gap between research and IT (e.g. data stewards).
• Enable the annotation of data already from the early beginning of the data creation process.
• Communicate data and software standards across communities and disciplines.
• Verify DMP and SMP throughout the research process.
• Establish incentives instead of formal requirements only (e.g. FAIR metrics, credits, review for data publication).

What are currently the biggest challenges for the development of FAIR Software/Services/Infrastructures?

General consideration: Management of requirements and expectations is one of the biggest challenges for the development of FAIR solutions.

• A clear definition of FAIR software/services/infrastructures for research data management is needed. This will allow the creation of appropriate evaluation criteria.
• Data diversity and community management are a challenge.
• Data validity as well as clean and continuous documentation throughout the research process is an issue.
• There is a lack of best practices, tools and applications.
• There is a lack of accepted impact and citation systems.
• Professionally skilled staff is urgently required. (> GO TRAIN)
• Legal issues, such as licensing information and options are a hurdle.
• Financial resources and funding for the development of FAIR solutions are needed.

What kind of support is needed in order to foster cultural change?

General consideration: Different communities – different paces!

1. Disciplinary associations are important agents to support cultural change. (>> GO BUILD)
2. Incentives to support cultural change are:
   • Making FAIR handling of data (and other digitial research objects – e.g. software) a criterion for funding.
   • Making FAIR handling of data (and other digitial research objects – e.g. software) relevant for career advancement.
   • Making FAIR handling of data (and other digitial research objects – e.g. software) relevant for reputation gain.
3. Establishment of an “culture of error”.
   Errors that occur during the data lifecycle should be acknowledged as normal and not result in “shaming”. Errors could for example be reported via anonymous systems (such as “LabCIRS”). It was assumed that an error culture can decrease researchers’ scruples when it comes to sharing data.

What prevents you/your colleagues from providing your data in a FAIR manner?

General consideration: Positive examples of fair data sharing need to be spread to gain more attention.

1. Researchers do make their data available, but most often only to their peers. Making data available beyond their peer group is allegedly inhibited by:
   • Fear of misuse
   • Insecurity in terms of legal issues/liability/GDPR
   • The wish to keep one’s one data for “later”
   • A lack of incentives
   • A lack of funding
   • A lack of infrastructure
   • A lack of institutional support
   • A lack of curation expertise

Which advantages do you hope for when ensuring FAIR compliance?

General consideration: Advantages are ranging from immediate advantages for the researchers to advantages for the society as a whole.

1. Immediate advantages:
   • Encourages quality insurance of data
   • Supports efficient research by fostering interdisciplinarity and enhancing cooperation opportunities. Machine readability and linking of data may lead to a usage above its immediate purpose.
   • Citing increases the researcher’s impact and adds to the reputation.
   • Preserves contextual knowledge on the data (interoperability).
   • Enables to re-use own data
   • Receives funding for FAIRification efforts.
   • Open licenses and transparent rules decrease bureaucracy.
2. Advantages for the societal and science system
   • Endorses transparency
   • Fosters citizens’ trust towards science

Who needs to be trained?

• Young talents, such as students and research assistants, need to be trained as they will shape the science system in the future and can expeditiously bring change and innovation. (>> GO CHANGE)
• Trainers and lecturers in research data management need to be trained.
• Professionals need to be trained. This includes researchers, lecturers, persons from the management level as well as from industry, evaluators and reviewers, future curators and colleagues from the library sector.
• Further target groups are computer scientists who develop infrastructure and services.
• Legal experts belong to the wider target group, too.
• Policy makers and scientific organizations must be involved in the training to develop an understanding of FAIR and initiate processes. (>> GO CHANGE)

What is the essential content of a training?

1. Within the organizational framework information needs to be given about:
   • Good science practice (>> GO CHANGE)
   • Policies, strategies and sustainability efforts (>> GO CHANGE)
   • Funding opportunities (>> GO CHANGE)
   • (Inter)national landscape
   • FAIR infrastructure and service provider
   • Best practices
   • Consulting services
2. Content aspects of training covers
   • Definition of terms
   • Data Management Plans
   • Software Management Plans
   • Data curation
   • Relevant (meta) data standards
3. Technological aspects of a training addresses (>> GO BUILD)
   • Repository systems, federated infrastructures and services (Validation)
   • Storage of data and version
   • Interfaces and interface development
   • Machine readability, interoperability and semantic search
4. Legal aspects include information about
   • Licenses of data for reusability aspects
   • Legal certainty for data re-use
5. Ethical aspects addresses
   • Protection of data from a scientific perspective (e.g. Who owns the data?)
6. Own use cases examples
   1. Working with and preparing own data

What is the best way to convey FAIR skills to specific target groups?

General consideration: Create advisory services as well as general and discipline-specific training for different target groups.

1. Install a central point of contact to give continuous impetus (>> GO CHANGE)
   • Offer constant support and advice (e.g. consultation hours)
   • Promote activities in research data management and enable networking
   • Collect and spread best practices
2. Train the management level and research data management experts by (>> GO TRAIN)
   • Providing advice and consulting.
   • Including FAIRness in online tutorials on good scientific practice.
3. Scientists and students are another target group of importance:
   • Integrate FAIR skills into the curriculum.
   • Offer online tutorials and webinars on good scientific practice.
   • Organize workshops, meet-ups and “bring your own data”- hackathons.
   • Offer discipline-specific training sessions.