What to expect

• Short, focused talks (10–15 minutes): Lightning-style presentations from academics, PDRAs and PhD students on how AI is enabling scientific research and discovery.
• What worked (and what didn’t): Each speaker will briefly cover (i) the core problem, (ii) the AI method(s) used, (iii) what was learned/achieved, and (iv) limitations—data issues, compute constraints, evaluation challenges, reproducibility, governance/ethics, etc.
• Poster + networking session (over lunch): A lively, informal opportunity to circulate, ask questions, and identify overlap across projects. Posters are strongly encouraged.
• Breakout discussions: Facilitated small-group sessions to identify shared needs (data, tools, compute, evaluation, governance), collaboration opportunities, and concrete next steps—particularly towards future external funding and cross-faculty initiatives.
• Practical takeaways: A curated set of themes, candidate collaborations, and an initial “opportunity map” for AI for Science activity at Lancaster.

Call for presenters (talks + posters)

We are actively seeking short talks (10–15 minutes) on any activity that fits “AI for Science” in the broadest sense. This includes methodological work, applied case studies, infrastructure/tooling, or evaluation and governance approaches that enable scientific progress.

Examples of suitable topics include (not limited to):

• AI for scientific discovery (new hypotheses, new materials, new mechanisms, new insights)
• AI for simulation and scientific computing (surrogates, emulators, PDE learning, accelerators)
• AI for experiments and measurement (automation, control, adaptive/active learning, lab/field workflows)
• AI for data assimilation and inverse problems (uncertainty quantification, calibration, causal inference)
• AI for earth and environmental science, climate, ecology, biodiversity, geoscience
• AI for health, medicine, life sciences, imaging, genomics, epidemiology
• AI for engineering, energy systems, manufacturing, sensors, robotics
• AI for evaluation and robustness in scientific settings (ground truth scarcity, dataset shift, validation)
• Responsible and trustworthy AI for scientific practice (reproducibility, data governance, ethics)

Poster displays: If you have a poster we will provide space for displays and discussion during the lunch session. We can also cover the cost of printing the poster.

Who should attend

• Academic staff and researchers working in, or curious about, AI-enabled science
• PDRAs and PhD students with AI-for-science methods or applications
• Research software engineers, data specialists, lab/technical teams, and professional services colleagues supporting research
• Anyone interested in building cross-disciplinary collaborations around scientific AI

Provisional schedule (10:00 AM – 3:00 PM)

• 10:00 – 10:15 Welcome and aims for the day (DSAIL)
• 10:15 – 10:45 Framing session: “What counts as AI for Science?” (shared language, examples, evaluation expectations)
• 10:45 – 12:15 Lightning talks (10–15 minutes each) + structured Q&A
• 12:15 – 1:15 Lunch + poster session + networking
• 1:15 – 2:15 Breakout sessions (theme-led groups; capture opportunities and blockers)
• 2:15 – 2:45 Report-back from groups (short summaries, candidate collaborations)
• 2:45 – 3:00 Close: Wrap-up and next steps.

Light refreshments will be available throughout the day. Lunch will also be provided, please indicate dietary requirements via the Eventbrite registration.

How to get involved

• Attend: Register for AI for Science Workshop
• Present: Submit a proposed title + 2–4 sentence summary with your registration
• Poster: Indicate in your registration that you would like to bring a poster (and whether you need DSAIL to pay the printing charge)