As one of the world's leading synchrotron facilities, MAX IV has always relied on advanced technology to enable groundbreaking science. Today, artificial intelligence (AI) represents a transformative technology with the potential to change how experiments are conducted, data is analysed, and scientific discoveries are made.
“AI is strategically important for MAX IV because our long-term scientific competitiveness depends on how effectively users can turn increasingly complex experiments and growing data volumes into new knowledge,” says Olof ‘Charlie’ Karis, Director of MAX IV. “I see three essential pillars for our AI strategy: managing and reducing data close to the instruments as modern detectors generate ever larger data streams; improving software and analysis workflows so users can make better and more efficient use of their data; and lowering barriers to access through tools such as an AI navigator, helping researchers navigate our capabilities and plan their experiments. Together, these developments can help us make fuller use of MAX IV and enable more researchers to tackle questions that are currently too demanding in terms of data, time or specialist expertise.”
AI’s potential in the Era of Data-Intensive Science
Modern synchrotron experiments generate big volumes of data. Advanced detectors produce information at unprecedented rates, while experimental techniques continue to increase in complexity and sophistication. As datasets grow larger and more multidimensional, traditional analysis approaches are becoming increasingly challenging.
AI offers new ways to address this challenge. Machine learning algorithms can identify patterns, detect anomalies, classify data, and reveal relationships that may otherwise remain hidden. By applying AI methods throughout the scientific workflow, researchers can analyse data faster, explore more complex questions, and gain new insights from their experiments.
Mirjam Lindberg, Interim Technical Director at MAX IV, sees potential: “New possibilities arise every day, and user expectations are rising. Keeping up with developments on many different fronts is a real challenge. At the same time, one of our greatest opportunities is the expertise and drive within our own user community. By working closely with AI-savvy researchers who are ready to push the boundaries, we can explore and validate new approaches in real scientific workflows. The challenge for MAX IV is then to turn what we learn with these early adopters into robust and accessible capabilities that benefit the broader user community, including researchers who are not AI experts themselves.”

This highlights an important principle behind MAX IV’s approach to AI: innovation begins with pioneering users, but long-term success depends on transforming promising ideas into robust capabilities that can be used across the wider scientific community.
“As a user facility, MAX IV continuously supports the current and future research needs of its academic and industrial user community and, by doing so, is an essential part of the long-term success of research and innovation in Sweden and beyond,” says Lindberg. “For AI to become part of this scientific journey, we need to rely on a solid foundation, with a robust data management infrastructure, modern, reliable experimental control, and well-understood and optimized scientific workflows. Our goal is to integrate AI capabilities at MAX IV in a way that enhances our beamline capabilities and benefits our user communities. This means that we must secure the foundation first because if the data, experiment setups, and workflows are not reliable and well connected, AI will remain a collection of interesting experiments rather than becoming part of how we deliver science.”
This emphasis on infrastructure aligns closely with MAX IV’s broader vision for AI. The ambition is not simply to deploy new digital tools, but to create an environment where AI can become a natural part of scientific research. MAX IV is already exploring practical applications of the new technology through pilot projects based on experimental data from its beamlines. One example uses data from HIPPIE beamline to investigate advanced machine learning techniques, including latent-space exploration of complex spectroscopic datasets. The objective is to develop methods capable of identifying meaningful structures, clusters and anomalies within large volumes of scientific data.
Collaborating with the research ecosystem
A key element in developing AI capabilities is collaboration with the research infrastructure ecosystem. Through partnerships with the National Academic Infrastructure for Supercomputing in Sweden (NAISS) and The National infrastructure for Data Visualization (InfraVis), MAX IV aims to provide researchers with access to advanced computing resources, modern visualization environments, and AI-enabled analysis tools.
“Sweden already has excellent national capabilities, such as the new Arrhenius supercomputer at NAISS, as well as world-class expertise in areas such as advanced visualization, as provided by InfraVis,” says Lindberg. “Rather than replicating those capabilities at MAX IV, we aim to integrate and streamline them into our beamlines, thus supporting and accelerating the scientific and innovation ambitions of more than 2000 users at MAX IV annually. The collaboration with NAISS and InfraVis allows us to develop and test that operating model in practice. One of the most interesting challenges is designing an AI-enabled data and computing workflow across the entire chain: deciding what to process close to the experiment, which data to retain, what to move to larger computing resources, and where AI can most effectively turn that data into scientific insight. MAX IV needs to be very strong where the experiment happens, but also very good at connecting our data and users to national capabilities in computing, analysis and visualization. By developing this, researchers will be able to move seamlessly from data collection at MAX IV to large-scale computing, analysis, and visualization.”
International collaboration
Since MAX IV is part of an international research community, collaborations also encompass actors outside of Sweden. Lindberg explains:
“The Genesis Mission in the United States is an interesting example of resources and expertise being brought together and coordinated among all actors funded by the Department of Energy (DOE) and beyond to dramatically accelerate AI-enabled science. For MAX IV, developments like this can provide both inspiration and opportunities for collaboration.”
A recent example is the collaboration between MAX IV and the Advanced Light Source (ALS) at Lawrence Berkeley National Laboratory. Through joint workshops and technical exchanges, experts are exploring how AI can enhance data analysis, experimental workflows, and scientific computing environments. The collaboration also provides an opportunity to explore what an AI-ready research facility should look like and how large-scale scientific infrastructures can evolve alongside rapidly advancing AI technologies.
“Being AI-ready is not about using AI everywhere,” says Lindberg. “It is about having the data, computing, software, and scientific structures in place so that AI can be introduced where it genuinely adds value. Our collaboration with ALS is particularly interesting because we are approaching the same challenge from different angles. MAX IV has ten years of experience as the world’s first fourth-generation lightsource, while ALS is now preparing for its major upgrade, which will bring new experimental capabilities and increasingly demanding data and computing requirements. That creates a very concrete opportunity to learn from each other. We explored exactly these questions in our bilateral MAX IV-ALS workshop in August, bringing together expertise in controls, data acquisition, scientific computing, data management and AI-enabled workflows. The aim is not simply to exchange ideas, but to identify areas where we can develop approaches together and learn from each other’s experience. This also connects to what I find interesting about the Genesis Mission: the focus on connecting experimental facilities, data, computing, and AI capabilities into a larger ecosystem. Through ALS and Berkeley Lab, we have an opportunity to be part of these conversations and consider what a similarly coordinated approach could mean in a Swedish and European context.”
The collaboration highlights a shared ambition: to create integrated ecosystems where AI becomes a natural component of the scientific process, supporting researchers from data acquisition through analysis and interpretation.
The vision is clear, MAX IV will be a research environment where advanced experimental capabilities are seamlessly combined with intelligent digital support, empowering scientists at every stage of the research process. Through strong national partnerships, international collaboration and a commitment to user-focused innovation, the facility is laying the foundations for an AI-enabled future in which more researchers can tackle increasingly complex scientific challenges and move more rapidly from experiment to understanding.
Importantly, AI is intended to complement and support human expertise, not replace it. Scientific creativity, interpretation and critical thinking remain central to the research process. AI provides tools which allow researchers and technical experts to work more efficiently and focus more of their time on discovery and excellence.