Since 2021, the SDSC has run an expert hub at the Paul Scherrer Institute to support the adoption of data science, machine learning and AI at Switzerland’s largest research institute and home of many large-scale scientific infrastructures (LSI) – such as the Swiss Free Electron Laser (SwissFEL) or the Swiss Light Source (SLS), where experiments conducted at these facilities can generate up to hundreds of terabytes of data per day.
Large-scale scientific infrastuctures bear a unique capacity for groundbreaking research and innovation. Through its collaborative praojects and community events like this workshop, the SDSC contributes to exploring the full potential of large-scale infrastructures by bringing modern data science and AI into their routine operations, turning massive data streams into actionable knowledge that drives technological development.
SDSC presentations included:
- Introduction to the SDSC – Prof. Olivier Verscheure, Executive Director, SDSC
- Data science for large scale infrastructures – Dr. Benjamin Bejar Haro, Large-Scale Infrastructures Domain Lead, SDSC
- Software Engineering and SDSC’s Renku Platform – Dr. Rok Roskar, Head of Platforms, SDSC
- Don't Just Read the Docs, Run Them! Imaging Plaza: a multimodal agent for distributed image analysis – Carlos Vivar Ríos, Senior Open Research Data Engineer, SDSC
Community presentations included:
- Current status and future activities of AI usage and organisation at PSI – Dr. Alun Ashton, Head of the Science IT Infrastructure and Services Department, PSI
- Current and future onsite computational resources in support of the Large Scale Facilities at PSI – Dr. Leonardo Sala, Science IT Infrastructure and Services, PSI
- Bridging simulation and experiment at PSI: reproducible AiiDA workflows and integration with Renku – Dr. Julian Geiger, Laboratory for Materials Simulations, PSI
Data science project challenges presented included:
- SwissFEL and other Infrastructures: Closing the MLOps gap – Dr. Christopher Arrell, Laboratory for Condensed Matter, PSI
- MX Detectors: Develop self-consistent diffraction model optimizing existing pipeline – Dr. Filip Leonarski, Beamline Data Scientist, PSI
- Multimodal imaging: Develop automated workflow for better data analysis – Dr. Elisabeth Müller, Laboratory for Multiscale Bioimaging, PSI
- Play It Again: Recovering analogue audio-video data with synchrotron in integrated analysis pipeline – Dr. Sebastian Gliga, Laboratory for Condensed Matter, PSI
- From Data to Answer at the Beamline: Analysis-driven acquisition and retention – Dr. Andreas Menzel, Laboratory for Macromolecules and Bioimaging, PSI
- End-2-End self-driving imaging facility: Imaging Plaza meets Beamlines – Dr. Goran Lovric, Laboratory for Macromolecules and Bioimaging, PSI
- ML-driven material development based on on-the-fly beamline data analysis – Dr. Grigory Smolentsev, Laboratory for Catalysis and Sustainable Chemistry, PSI
About the Swiss Data Science Center (SDSC)
The Swiss Data Science Center (SDSC) is a national research infrastructure in data science, machine learning and artificial intelligence (AI), founded by EPFL and ETH Zurich. Its mission - to enable data-driven science and innovation for societal impact - drives its initiatives in research projects, knowledge and technology transfer, and education. With a large multidisciplinary team of professionals in Lausanne, Zurich and at PSI in Villigen, the SDSC provides expertise and services to various domains, such as health and biomedical, energy and sustainability, climate and environment, digital society, and large-scale scientific infrastructures. The SDSC also contributes to initial and executive education programs at EPFL and ETH Zurich. For more information please visit: www.datascience.ch