Veranstaltungsprogramm

Abstract: The ongoing digital transformation profoundly affects scientific research, altering workflows, the nature of scientific artefacts, and public trust in science. Research processes have shifted towards increased reliance on digital tools, data-driven methods, and AI technologies. These innovations enhance efficiency and reproducibility but introduce complexities regarding data integrity, security, and ethical considerations. For instance - just to name some examples -, tools like Laboratory Information Management Systems (LIMS) and Electronic Lab Notebooks (ELNs) streamline data acquisition and management, creating ‘Smart Labs’ that offer greater precision and ease of collaboration across disciplines. In biomedical simulation and engineering, AI-driven methods allow for more intricate, scalable problem-solving. ‘Protected Data Spaces’ enable secure data-sharing methods for sensitive data, particularly in sports and health science. The transformation of scientific publishing, particularly how large datasets are integrated into new publishing models, have important use cases in climate research and other scientific communities dealing with big datasets.

However, digitalisation also raises questions about the nature of scientific results. With AI and machine learning algorithms generating insights, the traditional pathways of scientific discovery are evolving. The emphasis shifts from hypothesis-driven research to model-based predictions, challenging conventional notions of scientific validation. Moreover, large datasets and complex simulations are now integral parts of scientific publications, requiring new methods for data storage, curation, and peer review. The transition to these new publication formats prompts the need for open-access, FAIR-aligned infrastructures that ensure data is Findable, Accessible, Interoperable, and Reusable. The recently established Leibniz Science Campus “Digital Transformation of Research” (DiTraRe) investigates these shifts through interdisciplinary collaboration, developing new standards for data protection and open science.

The digitalisation of research not only affects how science is conducted but also how it is perceived by the public. Transparency, data openness, and the potential for reproducibility enhance public trust. However, the reliance on black-box AI systems and large-scale datasets raises concerns about the reliability and accessibility of scientific findings for the general public. This trust issue can be mitigated through improved science communication, emphasising the value of citizen science and engaging science-society intermediaries. The integration of digital methods into research workflows also invites more participation from non-experts, contributing to the democratisation of science. DiTraRe's work focuses on addressing these challenges by developing frameworks that balance innovation with the need for transparency, ethical standards, and public trust. By systematically analysing these issues, DiTraRe seeks to provide sustainable solutions that promote trust in scientific processes and outputs, ensuring that science continues to serve society in the digital age.

In today's changing world, developing robust research data management (RDM) services requires agile practices that adapt to shifting demands and increasing responsibilities. During the development of Coscine — a platform aimed at facilitating the FAIR principles — we are confronted with two major changes: 1) a steadily growing number of stakeholders and 2) constantly evolving RDM concepts and technologies. In our presentation, we show our journey to serve both — our stakeholders and development needs.

At the beginning of Coscine's development as an open-source platform in 2018, the development team worked with a startup mentality and focused on the rapid development of functional prototypes for RDM. The toolchain was featuring simpleArchive, a web based archive tool that allowed metadata management and PID assignment. Funded by the first stakeholder RWTH Aachen University, the team started to work with methods from Scrum and Kanban in GitLab —managing requirements on issue boards. While transferring the simpleArchive to an RDM-tool for the active project phase, two doctoral theses in RDM and computer science were involved in the development of Coscine and led to a high degree of research orientation and innovation.

From 2020 till 2023 Coscine was in pilot phase, which allowed to test various functions and understand the key needs of users - access to storage space in a FAIR environment for “warm” research data. Due to the integrative character of the platform Coscine was able to participate and accommodate in successful projects (e.g. VVMethoden, AIMS, NFDI4ING, NFDI-MatWerk, Coscine.nrw).

To effectively orchestrate the stakeholders, the management of their requirements in a transparent, open, and fair way became more intricate. Thus, we changed our workflow and incorporated roles and elements from the Scaled Agile Framework to improve prioritization and workflow management. Requirements were collected in public epics in GitLab, allowing stakeholders to track progress of their individual project demands. To define the order of epic development, priorities were set by the Weighted Shortest Job First (WSJF) method.

As Coscine entered regular operation in 2023, we were confronted again with a change – balance the adding of new features while maintaining existing services. The growing user base demanded enhanced stability and reliability. Implementing quality checks and high-availability practices became essential, requiring not just new processes, but also a cultural change within the development team. We reevaluated our prioritization methods, moving away from approaches like WSJF (focus on shortest jobs) to user-value, time criticality and risk reduction. This shift acknowledged that long-term topics require sustained focus and that balancing feature development with maintenance tasks is crucial for the platform's longevity.

Throughout our journey with Coscine, we've learned the importance of adaptability and continuous workflow improvement. This demands a great deal of communication and flexibility from the whole team but is rewarded by satisfied stakeholders and an up-to-date RDM service. While we've made significant strides in aligning our agile development practices with the demands of a changing research world, we know that we must keep on adapting.