Parallel Sessions
Per starting time | Per room | Session contents
Per starting time
Per room
Alphabetical, with summary
A deep dive into foundation models
Daan Korporaal
In this session we will go into how foundation models in AI are trained, giving some theoretical background and examples, as well as some practical examples from our own work. A more in depth session after last year's success.
13:55-14:40 in Plenary room
Theme: AI
Aanzet Agenda Rekenmodellen Digitale Tweelingen
Wies Vullings & Jene van der Heide
Nederland staat voor complexe ruimtelijke opgaven, zoals klimaatadaptatie, woningbouw, energietransitie en mobiliteit. Om deze opgaven versneld en effectiever aan te pakken, is digitalisering van processen in het ruimtelijk domein onmisbaar. Centraal hierin staat het ontwikkelen, beheren en toepassen van Digitale Tweelingen met interoperabele rekenmodellen als motorblok. Deze Aanzet Agenda bevat vier samenhangende sporen waarlangs deze rekenmodellen kunnen zorgen voor het versnellen en verbeteren van ruimtelijke processen: (1) Transparantie en Overzicht; (2) Standardisatie en interoperabiliteit; (3) Organisatie en Financiën en (4) Delen van modelkennis. De agenda is opgesteld in opdracht van ministerie van BZK. Het is een vervolg op het rapport 'Verkenning interoperabiliteit van rekenmodellen en digitale tweelingen' van Geonovum en het beschrijft concrete stappen om het juist omgaan met rekenmodellen te stimuleren.
13:00-13:45 in Plenary room
Theme: Model management
Agro Data Services Layer, governance & research
TBD
The farmmaps working group is currently working on the Agro Data Service Layer. This has consequences with respect to governance and research. At the WMDD we would like to present this and discuss this with colleagues from WUR.
15:00-15:45 in Quantum 3
Theme: Infrastructure & IT
AI supported analysis of high throughput phenotyping in quinoa breeding combined with genomic prediction
Robert van Loo
I will show how in several connected projects data are collected of a large set of quinoa genotypes (genomic marker data, expression data, UAV data in field trials, high throughput phenotyping data in NPEC) and discuss opportunities and challenges in using such data to improve routine prediction of breeding values of F1-hybrids in quinoa and support choice of parental lines for making new F1-hybrids using (AI-supported or classical statistical) predictive modelling of the performance of parental lines in smaller subsets taking from all possible F1-hybrids.
15:00-15:45 in Quantum 1
Theme: AI
Ambivalence workshop: exploring value conflicts in futures of agriculture
Mareike Smolka & Bas Boom
Visions of “agriculture 4.0” depict technological systems with sensors, AI, and robotics that measure farming conditions, inform decision-making, and automate tasks. Social science research on the future of agriculture pitches promissory visions in media, innovation, and policy discourses against critiques of such visions by agricultural stakeholders. Based on qualitative research, we challenge this bifurcation by demonstrating that ambivalence emerges when diverse stakeholders imagine futures of agriculture. By attending to ambivalence, we recognize value conflicts. In this workshop, we discuss potential alignments between conflicting values in futures of agriculture inspired by research and development at OnePlanet research center.
09:30-10:15 in Quantum 4
Theme: Value creation & impact
Automatic data transfers
Nika Valencic Kool, Joris Luijsterburg, Bruno Freire dos Santos & Erik van den Bergh
Are you dealing with data from different sources, like sensors, W-drive, Teams, PC, and you would like to establish periodic and automatic data transfer to Yoda, iRODS, HPC,..? Help us choose the right platform to fit your needs! In this session we will present two possible solutions, Neptune and Airflow. There will be time to ask questions and discuss your specific needs with experts on this topic.
10:30-11:15 in Quantum 4
Theme: Infrastructure & IT
Climate data for impact: applications and user perspectives
Natalia Gomez, Spyros Paparrizos, Rutger Dankers & other presenters TBD
What climate information do you, or your users, need? As climate change becomes ever more prominent, this session will discuss how WUR researchers are using climate data to produce valuable information for stakeholders. We will demonstrate a few practical examples of how climate data are being used within WUR, followed by an interactive discussion on climate data and information needs and issues that you may encounter.
13:00-13:45 in Quantum 4
Theme: Value creation & impact
Co-Creation in Action: Better Tools Through opensource Collaboration
Tim van Daalen
Do you ever have the feeling that something should already exist? That you can’t possibly be the first person to run into this challenge? I did — and I found others who had solved parts of the puzzle.
Together, we transformed separate one-off solutions into generic open-source tools that anyone can use today. This is a story about collaboration, shared ownership, and how investing extra effort up front can pay off through better tools, broader adoption, and contributions from others.
10:30-11:15 in Quantum 1
Theme: Infrastructure & IT
Connecting the economic and bio-physical drivers of food security: Integrating the global and local scales through a multi-model collaboration.
Jason Levin-Koopman, Walter Rossi-Cervi, Feroz Islam & Marijn Gulpen
In this session we integrate two global models – MAGNET a macro-economic model with global overage and national level resolution, and LPJmL, a global gridded model of climate, hydrology and crop growth. We do this by means of the MAGNETGrid crop allocation model to explore the impacts of European food consumption on water stress in water scarce regions demonstrating the global to local linkages, connecting global markets for food demand with water stressed river basins around the world. We will also explore the possibility of including local to global feedbacks giving a fuller picture of climate vulnerabilities of European food supply.
13:00-13:45 in Momentum (1+2+3)
Theme: Value creation & impact
Data integration for modular product environmental impact assessment tool
Jeroen Weststrate, Quinta Bonekamp & Roel Helmes
WSER’s Modulair Agri Life Cycle Inventory Tool (MALCIT) performs automated environmental impact assessments in agri-food supply chains. Currently, inventory data from widely used sources (FAOstat, IFAstat, Water Footprint Network) are ingested and processed in environmental impact calculations through custom Python workflows. While effective, this approach is difficult to scale and maintain.
MALCIT is developing a robust, flexible, and centralized data management approach that improves traceability and quality control, and enables efficient integration of multiple datasets. We call for colleagues in WUR with technical expertise and similar needs to help build a scalable solution that supports future growth and more reliable environmental assessments.
09:30-10:15 in Quantum 2
Theme: Infrastructure & IT
Data Protection Impact Assessment (DPIA): Benefits and Necessity
Team Privacy and Information Security
When processing data involves risks for those people involved. When do you use a DPIA? Who has what responsibilities? What are the various roles during the process. And finally how do you fill in a DPIA? As Privacy Officers we take you through this entire process.
13:00-13:45 in Quantum 1
Theme: Data management
DigilabTK
Bas van Vossen
NA
10:30-11:15 in Quantum 2
Theme: Value creation & impact
FAIR Data Station
Brett Metcalfe & Jasper Koehorst
For data to be reusable requires data and associated metadata to have similar formats, use consistent, standardised, terminology and units, and have the same structure. Ensuring that data meets these requirements can be a burden for researchers and it can be easy to overlook mistakes. To make data curation and management easier the FAIR data station (https://fairds.fairbydesign.nl/) has been developed that allows users to select appropriate metadata, generate a workbook, that can be filled in and validated ensuring high quality metadata. This hands-on session will introduce new users to FAIR Data Station (FAIR-DS) and update existing users to new features.
15:00-15:45 in Quantum 4
Theme: Data management
FairGraph: Exploring and Managing Integrated Research Metadata
Joeri Kalter (WUR) & Frans van der Kloet (AMC)
This session introduces FairGraph, an online platform for searching, exploring, and managing integrated research metadata across clinical, genomic, and other linked databases. Participants will learn how the platform supports natural-language dataset search, open-ended browsing, visual topic discovery through word clouds, and dataset organization through its data management features. The presentation highlights how FairGraph helps users navigate complex research metadata, identify relevant resources, and maintain structured, accessible datasets.
09:30-10:15 in Quantum 1
Theme: Data management
From Research-software-idea to a reliable service: How to Get Your Software Managed and Monitored
Research IT Solutions (RITS) team (t.b.d.)
Many researchers create software, but turning it into a reliable, sustainable service is challenging. In this session, Research IT Solutions (RITS) presents a practical roadmap to guide software from early idea to monitored and managed service. Using real examples, we will show how structured approaches to deployment, monitoring, and maintenance can improve reliability and reduce technical burden. This session is ideal for researchers unfamiliar with research software engineering practices. We also invite discussion to gather feedback and explore how we can better support the research community.
09:30-10:15 in Quantum 3
Theme: Software management
Gender & Intersectionality Gaps in AI in AgriFood
Kelly Rijswijk & Mark Ryan
This workshop aims to raise awareness about gender and intersectionality issues when it comes to AI in agrifood. Through the creation of a co-learning and reflexive space people can identify their own biases and jointly identify research gaps. In this workshop we welcome participants working on (ethical and social aspects of) AI and digital agriculture. The core of the workshop will be centred around a few relevant AI in agrifood cases. Participants gets asked a number of reflective questions that allow them to explore ELSA aspects of the particular case and will be challenged to jointly develop research gaps.
13:55-15:45 in Quantum 2
Theme: AI
Hands-on generative AI for research
Valerio Bianchi
Most of us meet AI through a chat box. This 45-minute session shows a more useful path: driving generative AI yourself, from a Google Colab notebook. Live, I demo four everyday research tasks with Google's Gemini: prompting for text, analysing images, summarising documents, and answering questions through a web-search tool I build on the spot. You then open the shared notebook and run it yourself, leaving with a working starting point for your own work. Python-based but beginner-friendly. Bring a laptop and a Google account. No prior AI experience needed.
15:00-15:45 in Momentum (1+2+3)
Theme: Generative AI
How do the neighbours do it? Model & data quality assurance across Dutch research institutes
TBD
Models and datasets increasingly underpin scientific and policy advice — so how do we keep them trustworthy? Guidelines abound, but the real test is putting quality assurance into practice. In this session, colleagues from other Dutch applied-research institutes share how they actually organise model and data QA — and above all, the challenges they hit and how they cope. After a brief look at the WUR approach, each guest gives their perspective, followed by an open discussion with the audience. The aim is a candid, practical exchange — including how QA must adapt to AI models — so we learn from one another.
10:30-11:15 in Momentum (1+2+3)
Theme: Model management
How to best annotate csv data for interoperability (just asking, no final answer)
Paul van Genuchten (and maybe others)
We've recently been exploring different approaches for annotating soil observation datasets so they can be more easily understood, reused, and integrated across systems. As part of the SoilWise Horizon Europa project, which aims to facilitate FAIR aspects in the Mission Soil program. We tried RO-Crate, CSVW, GeoPackage, each offering interesting strengths, trade-offs, and perspectives. In this session, Paul van Genuchten shares what worked well, where we struggled, and how these approaches compare from a practical, researcher-oriented perspective. More importantly, I'm hoping to tap into the collective expertise in the room: What alternatives am I missing?
09:30-10:15 in Quantum 1
Theme: Data management
Hybrid Modelling Approaches for Prediction and Optimization in Food and Biobased Research
Ruud van der Sman, Xuezhen Guo, Floor Boon & Taniya Kapoor
This session presents the results of the ongoing hybrid modelling KB project and related research within the Agrotechnology and Food Sciences Group (AFSG) at WUR. It covers physics-informed neural networks and control-theory-based robust reinforcement learning, in which domain knowledge is integrated with data-driven approaches. These hybrid modelling frameworks are applied to predict product properties and optimize processing operations in the food and biobased sectors. The results demonstrate the strong potential of combining mechanistic and data-driven models to improve prediction, optimization, and scientific understanding in these domains.
13:55-14:40 in Quantum 3
Theme: Model management
Interpreting Explainable AI in practice
Osman Mutlu & Zuzanna Fendor
We will present a live demo of explainable AI (particularly SHAP) that people can follow along interactively, while talking about how to correctly interpret different types of explanations we see in practice. The purpose of this interactive presentation will be to give a practical introduction to SHAP, focusing on how to use it and how to read the results. Finally, we will give a presentation on a real-world use-case with its outcomes and pain points.
15:00-15:45 in Plenary room
Theme: Generative AI
Modeling for Pandemic Preparedness: Research Software from WUR IDE for COVID-19 and Beyond
Büsra Atamer Balkan
Although we do not know when the next pandemic will hit, we can and should aim to be better prepared. In this session, we will present a set of simulation models developed by the WUR Infectious Disease Epidemiology group. These tools were developed around COVID-19, but they also help us to explore questions for future outbreaks. Our work connects different scales, from how people interact in indoor spaces to how infections spread across schools, households and wider populations. We will discuss modeling choices, data needs, stakeholder access to these models and their insights, and model reusability for future outbreak preparedness.
13:00-13:45 in Quantum 3
Theme: Model management
Opening up Greenhouse Data Platform
Stef Maree, Kevin Mildau
As researchers, we generate extremely valuable data through crop growth experiments. In this talk, we demonstrate our work focussing on lowering the barrier to collaborate, share and open up our greenhouse data to promote joint research and AI developments. We share our struggles, lessons learned and what it could mean for you.
10:30-11:15 in Plenary room
Theme: Data management
plmTune: a framework for midtraining and finetuning of protein language models
Aurin Vos
NA
13:00-13:45 in Quantum 3
Theme: Model management
Research Software Publishing
Lingtong Gai & Koen Bokhorst
Did you know that your research software can be published, assigned a DOI and cited just like a research paper? Join this hands-on workshop to learn how to prepare your research software for publication and make it easier to be FAIR. We will cover the following elements of software publishing practices: README file, License, citation, documentation, code and repositories.
09:30-10:15 in Quantum 3
Theme: Software management
Reviving Legacy Instruments: Open Hardware, Open Drivers, and Digital Sovereignty in Research Infrastructure
Vincent de Boer & Vittorio Saggiomo
Across WUR's labs, technically functional instruments sit idle because their proprietary software no longer runs on modern operating systems. This is not just an e-waste problem, it is a research infrastructure problem: working hardware, locked out of modern data pipelines. We show how open hardware principles and open-source frameworks like PyLabRobot allow us to analyse instrument communication protocols and build open, compliant drivers for legacy devices, restoring them to current workflows. This reduces dependency on proprietary vendor ecosystems and strengthens digital sovereignty in research infrastructure. Drawing on hands-on experience, we discuss what it takes to make hardware a transparent, sustainable, and secure part of the data pipeline.
13:00-13:45 in Quantum 2
Theme: Infrastructure & IT
Searching for data and data reuse
Koen Bokhorst & Ignacio Sancho
FAIR data focusses on the reusability of research data, but in some fields, the reuse of data is still limited. In this session, we will explore different strategies for searching for existing research data. Next to this, we will also evaluate the quality and FAIRness of the data and the conditions (licences) under which reuse is possible. We will end with a short data search exercise, bring your own laptop.
10:30-11:15 in Quantum 3
Theme: Data management
Strategic modelling infrastructure
Not sure yet, but probably modellers from WPR, WSER, WLR, WENR and more
WR has been recognized as host of a model and data infrastructure on life sciences supporting policies and decision-making on national and regional level. We would like to share recent developments on future proofing this strategic modelling infrastructure.
09:30-10:15 in Plenary room
Theme: Model management
The Anunna Road: What do you need?
TBD
We will have an interactive Anunna roadmap session where we invite attendees to bring their ideas to the table so that we can actually incorporate them into our real roadmap.
13:55-14:40 in Quantum 4
Theme: Infrastructure & IT
The Digital Future Farm: Where data assimilation meets farm-level impact
Claudia Kamphuis, Frits van Evert, TBD
With the Digital Future Farm (DFF) we promise to bridge the gap between complex models and real-world decisions: is it possible? In this session, the DFF project shows how data assimilation, sensor integration, and crop models come together in a working farm environment. We briefly unpack the technical architecture and then give the floor to a practitioner from (De Marke/VDBORNE) who shares what the system actually changes on the ground. Which decisions it informs, what surprised us, where the gap remains. Closing off with an open discussion to see if digital twins are worth the investment.
09:30-10:15 in Momentum (1+2+3)
Theme: Value creation & Impact
The future of Data & Compute @ WUR.
NA
13:55-14:40 in Quantum 4
Theme: infrastructure & IT
The Invisible Work Behind AI: Building Reproducible Pipelines for Animal Behaviour Research
Rodania Bekhit
NA
13:55-14:40 in Quantum 1
Theme: AI
VibeCode craft session: making Data Visualizations (dashboards) using Gen AI
Eva-Tsvetalina Christoff
utilizing free GEN AI (copilot) to visualize our data: making creative dashboards, visuals, graphs.
how to prompt efficiently, NOT give our data to the cloud and create useful visuals for our scientific findings for cheap (short time). Participants will be able to make their own creative dashboard in 45 min
13:55-14:40 in Momentum (1+2+3)
Theme: Generative AI
WUR Model & Data Day 2026
Registration website for WUR Model & Data Day 2026WUR Events Teamdata@wur.nl
WUR Events Teamdata@wur.nlhttps://event.wur.nl/wur-model-and-data-day-2026/registration
2026-10-15
2026-10-15
OfflineEventAttendanceMode
EventScheduled
WUR Model & Data Day 2026WUR Model & Data Day 20260.00EUROnlineOnly2019-01-01T00:00:00Z
Akoesticum Akoesticum Nieuwe Kazernelaan 4 - D 6711 JC E Netherlands
