English

Towards Advancing Research with Workflows: A perspective from the Workflows Community Summit -- Amsterdam, 2025

Distributed, Parallel, and Cluster Computing 2026-02-06 v1

Abstract

Scientific workflows have become essential for orchestrating complex computational processes across distributed resources, managing large datasets, and ensuring reproducibility in modern research. The Workflows Community Summit 2025, held in Amsterdam on June 6th, 2025, convened international experts to examine emerging challenges and opportunities in this domain. Participants identified key barriers to workflow adoption, including tensions between system generality and domain-specific utility, concerns over long-term sustainability of workflow systems and services, insufficient recognition for those who develop and maintain reproducible workflows, and gaps in standardization, funding, training, and cross-disciplinary collaboration. To address these challenges, the summit proposed action lines spanning technology, policy, and community dimensions: shifting evaluation metrics from raw computational performance toward measuring genuine scientific impact; formalizing workflow patterns and community-driven benchmarks to improve transparency, reproducibility, and usability; cultivating a cohesive international workflows community that engages funding bodies and research stakeholders; and investing in human capital through dedicated workflow engineering roles, career pathways, and integration of workflow concepts into educational curricula and long-term training initiatives. This document presents the summit's findings, beginning with an overview of the current computing ecosystem and the rationale for workflow-centric approaches, followed by a discussion of identified challenges and recommended action lines for advancing scientific discovery through workflows.

Keywords

Cite

@article{arxiv.2602.05131,
  title  = {Towards Advancing Research with Workflows: A perspective from the Workflows Community Summit -- Amsterdam, 2025},
  author = {Irene Bonati and Silvina Caino-Lores and Tainã Coleman and Sagar Dolas and Sandro Fiore and Venkatesh Kannan and Marco Verdicchio and Sean R. Wilkinson and Rafael Ferreira da Silva},
  journal= {arXiv preprint arXiv:2602.05131},
  year   = {2026}
}
R2 v1 2026-07-01T09:36:57.581Z