NOBUGS 2026

AI-assisted coding is gaining significant traction at the EuXFEL. Among all LLM use cases facilitated by the facility’s enterprise foundation model provider, coding-related tasks have the highest token usage. However, there are still unanswered questions and evolving guidelines regarding AI-assisted and agentic coding. These include determining where we should embrace, tolerate, and avoid...

At SPring-8, scheduling accepted user proposals has traditionally required extensive manual coordination by beamline staff, who must consider user availability, staff availability, beamline operation cycles, and compatibility of experimental setups. To address this challenge, we developed an automated proposal scheduling system based on 0-1 integer programming and deployed it for practical...

The experiment data acquisition and sequencing system BLISS was introduced on the first ESRF beamlines in 2020 and is today, in 2026, driving experiments on all 46 ESRF beamlines. Data acquisition sequences are implemented for more than 100 scientific techniques.
BLISS is programmed in Python, to allow easy sequence programming for scientists and easy integration of scientific software. ...

Silx (https://github.com/silx-kit/silx) is an open-source Python toolkit and tool (silx view) for viewing HDF5/Nexus files used in many sites world-wide.

silx provides not only a NeXus-HDF5 data visualization tool but also a comprehensive collection of Python packages to support the development of data assessment, reduction, and analysis applications at synchrotron radiation facilities...

For over two decades, the dCache project has provided open-source software to meet ever more demanding storage requirements.
More than 80 sites worldwide rely on dCache to provide storage service for LHC experiments, Belle-II, EuXFEL, and many others.
This can be achieved only with a well-established software development process, from a whiteboard, where ideas are created, through...

Instrument scientists and scientific users desire a unified, flexible and accessible interface to control their experiments: one place from which to select the sample, compose and run scripted scan sequences, and view and plot live data with the possibility for multi-user sessions and remote access. The technical staff managing the control system, on the other hand, will want support for a...

Darfix is a GUI software to treat Dark-field X-ray Microscopy (DFXM) data at the ID03 beamline at ESRF.

Since its inception in 2019, Darfix specifications have evolved through successive revisions; a major change was the upgrade of the data acquisition format from EDF to HDF5. As a result, the Darfix codebase began to grow and became overly complex. An incomplete abstraction attempted to...

RESEDA (Resonance Spin Echo for Diverse Applications) is a spin-echo spectrometer at the MLZ and offers sub-µeV energy resolution and an exceptional dynamic range of ~8 orders of magnitude. One of its operation modes is the MIEZE (Modulation of IntEnsity with Zero Effort) mode, which enables the measurement of the intermediate scattering function, S(Q, τ), in depolarizing sample environments,...

With the upgrade from SLS to SLS 2.0, a fourth-generation synchrotron, the macromolecular crystallography (MX) beamlines have undergone significant hardware and software enhancements, enabling fully automated data collection workflows. By leveraging the on-the-fly data processing capabilities of Jungfraujoch (Leonarski, F. et al., 2023), together with integrated machine learning and artificial...

The National Synchrotron Radiation Centre SOLARIS, a third-generation light source, is the only synchrotron facility in Central-Eastern Europe, located in Poland. The SOLARIS Centre, equipped with seven fully operational beamlines and three under construction, serves as a key research hub for a wide array of scientific disciplines. The Centre requires advanced software tools to support the...

Knowing your data is trustworthy brings confidence and peace of mind to the subsequent data analysis step(s). This is reliant on the provision of thorough, well-configured data correction steps. Over the last decade, we have achieved this through implementation of a modular data correction system, which allows a comprehensive data correction graph to be constructed for a given instrument or...

The High Energy Photon Source (HEPS) generates massive amounts of highly heterogeneous experimental data, placing significant pressure on storage, computing, and network resources. Data compression is an effective way to mitigate data growth, but a single compression method cannot meet the diverse requirements in the HEPS context. To address this issue, this work first studies a compression...

Ptychography is a lensless, high-resolution advanced imaging technique widely used at modern synchrotron facilities. The integration of deep learning methods has partially addressed the issues of high computational cost and strict data overlap requirements associated with traditional algorithms. However, existing discriminative learning models still face limitations in reconstruction accuracy...

Estimating the latent capacity required to model high-dimensional data is important for representation learning, compression, model design, and downstream control. We formulate latent capacity estimation as a sequential measurement problem over an ordered family of nested representations induced by prefix-masked autoencoders. This turns latent capacity selection from a costly architectural...

Many existing scientific workflows for segmentation-based data acquisition rely on intensive manual labour of domain experts to label samples as out-of-box Machine Learning models do not generalise well. Different frameworks adopt different data formats, evaluation metrics, and configuration interfaces, resulting in slower experiment setups and need for intermediary steps for result...

Two-dimensional pixel detectors used at X-ray diffraction beamlines
provide planar cuts through the reciprocal-space representation of the
X-ray scattering function. With the advent of today’s high-brilliance
X-ray sources, the resulting very high signal-to-noise data demand
analysis pipelines that minimize both systematic errors and approximation
artifacts.

We present a new...

Modern beamlines need a flexible way to run everything from simple motor moves to multi-dimensional, hardware-triggered acquisitions - often orchestrated through web-based user interfaces rather than bespoke scripts. At the Australian Synchrotron (ANSTO), we have implemented a Generic Scan library on top of Bluesky, making Bluesky declarative: it specifies what should be measured, not how the...

At the Australian Synchrotron (ANSTO), we see the future of beamline control in increased automation and composable workflows: modular building blocks that can be rearranged as science use cases evolve, and as ML/AI-driven decision points become part of routine experiments. We have adopted Prefect as a workflow engine to orchestrate end-to-end beamline pipelines in a single system, spanning...

Bridging the gap between a visiting user's X-ray Photon Correlation Spectroscopy (XPCS) experiment concept and its technical feasibility at Argonne National Laboratory's Advanced Photon Source (APS) often requires extensive consultation with beamline scientists. An integrated AI platform at APS Beamline 8-ID is introduced, consisting of multiple components. An agentic AI system autonomously...

As part of the upgrade project of the Swiss Light Source (SLS) to a fourth-generation synchrotron, a unified solution for controlling and orchestrating experiments has been developed. The Beamline and Experiment Control system (BEC) leverages community tools but embraces a service-oriented approach to facilitate fast and maintainable developments in a rapidly evolving scientific...

Feedback from real users who rely on software daily is one of the most valuable inputs for developers. In scientific facilities using distributed control systems, large-scale GUIs must handle hundreds of process variables event notifications and polling requests, often affected by hardware delays and timeouts. Users expect reliable, fast data delivery, and startup time is the first hurdle:...

Large-scale research facilities such as the European XFEL rely on precise real-time tuning of numerous interdependent subsystems to ensure stable and optimal operation. Automating these procedures can significantly reduce the workload of operators while improving the efficiency and reliability of beamtime usage. In this work, we will present ongoing efforts aimed at automating various elements...

After 31 years of operations the Italian third generation synchrotron light source Elettra will be replaced by Elettra 2.0, a fourth-generation one. The upgrade project includes the construction of at least five new beamlines, bringing the total to 32. We are using the dark period to replace the existing beamline Personnel Safety System with the new Beamline Access Control System BACS 2.0....

Operando synchrotron experiments require coordinated control of complex sample environments and beamline measurements over several days. These experiments are often run manually or through monolithic scripts, limiting flexibility and making live adaptation during beamtime difficult. Within the ROCK-IT project, we developed a Bluesky-based orchestration framework that represents experiments as...

High Energy Photon Source (HEPS) experiments generate large volumes of image sequences, whose wide dynamic range, non-negligible noise, and complex temporal redundancy make it difficult for existing lossless compression methods to reduce data volume effectively. We propose a two-stage lossless compression method based on the Mamba state space model. The original images are first transformed...

Operando spectroscopy plays a central role in catalysis research, but its full potential increasingly depends on efficient automation of experiments, control systems, and data handling. Within the project ROCK-IT (Remote, Operando Controlled, Knowledge-driven, and IT-based), we developed a framework to support automated in situ XAS measurements of catalytic samples at the CATACT beamline at...

As a fourth-generation synchrotron radiation light source, HEPS generates BCDI data that far exceeds that of third-generation sources. The reading, processing, and saving of these data have become extremely challenging. Daisy-BCDI is a high-performance computing software developed based on openClaw. With the backend connected to the Gemma 4:31B open-source model API key, users can perform data...

EasyScience (https://github.com/easyscience) is a framework initiated by the European Spallation Source (ESS) to unify and streamline data analysis across neutron scattering techniques. Built with Python and QML, EasyScience aims to accelerate the analysis step in the data processing workflow by providing friendly graphical interfaces for new users and scripting options via Jupyter notebooks...

The ESS data path for science instruments has a natural split into distinct paths for detector data and controls data. Four elements make up the detector data path (data path): Frontend electronics, readout master, event formation unit and Apache Kafka.

The data path consists of a combination of firmware (FPGA) in the first two elements and server software in the latter two elements. This...

BESSY II operates more than fifty beamlines supporting a broad range of experimental techniques across diverse scientific fields. With unique undulators, world-record photon energy resolution (e.g., <1 meV at 60 eV), femtosecond time resolution, and nanometer-scale spatial precision, the facility enables highly demanding experiments. These increasingly include operando and multimodal...

A 20-year-old codebase, an ancient and inefficient dependency, a weakness tangled up in 1.5 million lines of code, and the proper tools to finally restore order and performance.
The Mantid project is an international software collaboration to create a program suite for every need in neutron scattering data reduction workflows, first begun in 2007. Since that time, it has relied on the API...

The increasing data production rates at accelerator-based light sources necessitate the development and adoption of efficient data reduction and compression techniques [1]. Pre-storage data reduction, as well as various forms of on-the-fly processing systems [2], are becoming increasingly prevalent. Hardware-accelerated computing, particularly heterogeneous GPU–CPU–FPGA systems [2], offers an...

SciLog [1] is an electronic log notebook developed and deployed at the Paul Scherrer Institute (PSI). As part of the Open Research Data [2] project, we have further enhanced its integrations with systems within the PSI infrastructure, as well as its interoperability with other FAIR data management tools.

In the context of experiment data management, a logbook is usually linked to a...

Powder diffraction is a notoriously ill-posed inverse problem: the observed one-dimensional pattern entangles lattice parameters, space group symmetry, atomic arrangement, microstructural effects, and instrumental contributions, which classical pipelines address sequentially with limited reproducibility. We present CrystalAI, a machine learning framework for end-to-end crystal structure...

Effective UI/UX design is essential for scientific graphical user interfaces (GUIs), particularly in environments where complex workflows and time-critical tasks challenge usability. Applying user-centered design principles, accessibility standards, and continuous user feedback can significantly improve the usability and efficiency of experimental physics control systems.

This presentation...

SIRIUS, a 4th generation synchrotron light source operated by the Brazilian Synchrotron Light Laboratory (LNLS), is starting its second phase of beamline projects and will support the future integration of Orion, a new BSL-4 national laboratory under development. The increasing number of beamlines and experimental modalities has expanded the scale and complexity of scientific software...

As Serial Micro- CT moves toward routine, high-throughput imaging of large sample collections, the underlying DAQ software must support flexible orchestration, distributed processing, and seamless integration of online services. We present a decentralized software architecture that addresses these requirements through a stream-oriented design. In this framework, an “experiment” is...

With recent improvements to the efficiency of collecting single crystal diffuse scattering at synchrotron x-ray sources, large contiguous scattering volumes of diffraction data comprising 100GB can be collected in 20 minutes with high dynamic range and low backgrounds. The Python package, NXRefine, implements a complete data reduction workflow, from ingesting the data, orienting the single...

Amber fossils offer an unparalleled glimpse into ancient ecosystems, preserving extinct organisms in three dimensions with exceptional detail -from external morphology to internal structures. The AMBCAT project aims to publicise the studies of these rare specimens by centralizing high-resolution digital scans in a globally accessible platform. By democratizing access to these invaluable...

Metadata standardization at complex scientific facilities, such as synchrotron light sources, is a challenging task due to the diversity of instruments, sample types, experimental techniques, acquisition strategies, and scientific questions supported by these facilities. Current approaches for standardizing metadata storage in NeXus files commonly focus solely on the sample-data acquisition...

Scientific research publication abstracts are often highly advanced and primarily accessible to scientific communities. To extend accessibility and broaden the audience beyond this community, an audience-specific summarisation of abstracts called Tailored Abstract Translator (TAT) was implemented using open-source Large Language Models (LLMs).

This implementation makes scientific knowledge...

Light-source facilities now generate data at rates that overwhelm conventional storage paths — HEPS Phase-1 alone averages around 800 TB per day with peaks of 3.2 TB/s, while typical parallel storage sustains only tens of GB/s. Compression is essential, and the community has developed many specialized algorithms — lossless, hand-crafted lossy, and AI-based lossy — each with different...

SIRIUS, the 4th generation synchrotron light source operated by the Brazilian Synchrotron Light Laboratory (LNLS/CNPEM), currently hosts multiple beamlines with different experimental techniques and continuously evolving software demands. As beamline activities and data processing requirements grow in complexity, challenges related to interoperability, maintainability, software reuse and...