Dresden 2026 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz
AKPIK 6: AI Methods for Physics and Materials Science
AKPIK 6.3: Vortrag
Donnerstag, 12. März 2026, 17:15–17:30, BEY/0127
Support for self-driving labs within the NOMAD ecosystem — •Sarthak Kapoor1, Hampus Näsström1, Ahmed Ilyas1, Alvin N. Ladines1, Alexander Fuchs2, Joseph F. Rudzinski1, Lauri Himanen1, Sebastian Brückner1, José A. Márquez1, Martin Albrecht3, and FAIRmat Team1 — 1Physics Department and CSMB, Humboldt-Universität zu Berlin, Germany — 2Department Physik, FAU Erlangen-Nürnberg — 3Department Materials Science, IKZ Berlin
Self-driving laboratories (SDLs) rely on robust digitization, structuring, and analysis of experimental data. We present NOMAD [nomad-lab.eu] [1] as a comprehensive research data management and workflow ecosystem that addresses the challenges inherent to emerging SDLs. The NOMAD ecosystem supports direct interfacing with lab instruments and addresses the transformation of instrument outputs into machine-actionable formats, a key requirement in SDLs, through a flexible schema system that allows researchers to represent raw data as standardized entries based on community-developed or laboratory-specific definitions. NOMAD Actions provide a robust framework for defining, executing, and monitoring sophisticated analysis and decision-making SDL workflows, such as ML pipelines and Bayesian optimization strategies. Moreover, NOMAD's workflow storage framework facilitates detailed provenance tracking, along with tools for navigating workflow graphs. Together, these capabilities position NOMAD as a foundational toolkit for realizing scalable, reliable, and FAIR SDLs. [1] Scheidgen, M. et al., JOSS 8, 5388 (2023).
Keywords: Research data management; FAIR; NOMAD; Self-driving labs; Workflows