TEACH 5In-Person Event

29 Sept 2025, 12:00 → 1 Oct 2025, 14:15 Europe/Berlin

HIDA Hub

The fifth edition of the "Talking about Education Across Communities at Helmholtz" (TEACH) community conference will take place from 29 September to 1 October – live in Berlin!

The TEACH conference is a participant-driven community event that brings together individuals passionate about teaching. Teaching staff, educational researchers, training coordinators, and HR developers join forces to share methods, best practices, and experiences. The aim: to strengthen the teaching community across the Helmholtz Association and its partner institutions – because high-quality education is the foundation of high-quality research.

---

Program overview

This year’s program will offer a vibrant mix of contributions from both the Helmholtz education community (talks, workshops, and a barcamp session) and invited external experts (keynotes and workshops). TEACH 2025 will provide rich opportunities for peer-to-peer exchange and structured professional development.

To kick off the event, we’re excited to welcome Dr. Leah Sharp from the Technical University of Munich (Department: ProLehre – Media and Didactics), who will deliver a keynote on "Teaching in the age of AI: Enhancing student engagement with AI in Education"

Additionally, we oragized two half-day workshops:

• Workshop 1: Interactions, Tools, and Methods in Online Teaching
  (Dr. Patricia Wohner, Hanke Teachertraining)

• Workshop 2: Effective Use of Large Language Models (e.g. ChatGPT) in Teaching
  (Julian Kortendieck, HIDA)

The remainder of the program is shaped by you (Note: Call for Contributions" is now closed).

You can find the final program on the left hand side.

---

Registration

Registration will be open until 5 September 2025.

---

Date & Location

29 September – 1 October 2025
HIDA Hub
Anna-Louisa-Karsch-Straße 2
10178 Berlin

---

Fees

Participation is free of charge.

---

TEACH in the past

Here, you can find the event pages of the last four TEACH events. 

TEACH I    -   TEACH II    -    TEACH III   -   TEACH IV

 

---

                      

Monday 29 September

13:00 Arrival & Networking

1 Opening

Welcome to this year's TEACH!

Speakers: Dr Anna-Lisa Doering (HIDA), Fredo Erxleben (Helmholtz-Zentrum Dresden-Rossendorf), Ines Schmahl (Forschungszentrum Jülich), Mirl Trösch (Geomar Helmholtz Research Center)

2 Keynote: Teaching in the age of AI - Enhancing student engagement with AI in education

Education has always evolved in response to technological developments: calculators redefined math education, Wikipedia reshaped access to knowledge, Google transformed research, and the global pandemic reimagined classroom boundaries. Each shift challenged us to adapt how we teach and to reevaluate our values as educators. Now, generative AI presents a transformation of an entirely different scale. It is not just a tool, but a force reshaping how we work and how we think, permeating every discipline and profession.
In this talk, we explore what the growing influence of AI means for higher education, and how our teaching should evolve to foster the future workforce that will meet the needs of a rapidly changing society. To prepare students for a world where competencies like adaptability, critical thinking, and ethical judgement are more significant than ever, educators must reimagine their roles. Rather than resist change, we must guide it by teaching not just about AI, but with it. Through reflections on past revolutions in education and the unique nature of the current shift, strategies for fostering future-ready competencies will be considered. This is not only about technical fluency, but about nurturing students who can thrive in complexity, act with integrity, and shape the future alongside intelligent machines.

Speaker: Dr Lea Sharp (TUM)

Talks

3 OpenEdX for Helmholtz

The Helmholtz Cloud offers a wide variety of digital services for all members of the Helmholtz association. One of the newest of those services is the e-learning platform OpenEdX.

In this talks a brief introduction to the platform will be given, its features and workflows will be explained and we will give a quick introduction how you can use it to set up and run online courses and make most of the wide array of features it offers.

Speaker: Norbert Spyra

4 Beyond the Hype: A Four-Layer Framework for Understanding AI's Role in Academic Work

The integration of generative AI in higher education poses a central paradox: the very activities that develop critical thinking – literature searching, reading and writing itself – are precisely those being transformed by AI integration. This talk introduces a comprehensive four-layer framework developed from an information sciences and writing sciences perspective, which serves as the theoretical foundation for the online course "Writing and Information Retrieval with Generative AI" (https://llm-literacy.de/ki-at-helpbw/). This framework enables educators to systematically understand and address AI's impact on academic work while maintaining educational integrity.
The framework distinguishes four levels where AI can substitute academic work:

1. Form Layer: Language mechanics including grammar, style, and linguistic correctness. AI excels here, having been trained on vast datasets to produce grammatically correct sentences across languages.
2. Content Layer: Knowledge and meaning beyond linguistic form. AI operates indirectly at this level, generating synthetic text without genuine understanding or intentionality. While AI can reproduce knowledge, humans must recognize and validate the meaning within AI-generated content.
3. Process Layer: The complex cognitive processes of research, writing, and learning. AI simulates these processes through role-playing interactions but cannot genuinely experience critical thinking or knowledge construction.
4. Procedural Layer: AI deployment within research methodologies, either as support tools (e.g., laboratory documentation) or as research methods themselves. This layer is highly discipline-specific and requires context-dependent evaluation.

The talk demonstrates how each layer is taught and applied in different educational settings through concrete examples from the online course. This practical implementation of the framework enables students to navigate AI assistance across all four layers while sharpening critical evaluation skills for AI-generated content. Additionally, the framework functions as a guideline for developing appropriate examination formats and assessment strategies that account for AI integration in academic work.

Speaker: Kevin Schumacher (KIT Library and KIT Writing Center)

16:10 Ice breaker & coffee break

Parallel workshops & talks

5 The Educator’s Role in Designing AI-Assisted Training Materials: A Hands-On Workshop

The rise of generative AI offers exciting opportunities to streamline the creation of training materials, but it also raises important questions about pedagogical quality and the evolving role of human expertise.
In this hands-on workshop, we will explore what the role of educators is—and should be—when using AI tools like ChatGPT to create instructional content. Participants will be guided through the process of developing training materials both with and without generative AI, allowing for a direct comparison. Together, we will critically examine the advantages and limitations of AI-assisted design, discuss evaluation strategies, address ethical concerns such as plagiarism, and explore how to maintain creativity and originality. Combining practical activities with group reflection, the session aims to equip participants with responsible, human-centered approaches to using AI in education.

Speakers: Hamideh Haghiri (DKFZ), Dr OEzlem OEzkan (HMC)

6 Using Neo4j to Organize and Explore Educational Content

Neo4j is a global leader in graph database technology and an excellent tool for organizing and managing learning content. Its flexible and intuitive data model makes it particularly well-suited for educational contexts, where information such as courses, topics, prerequisites, and user progress is deeply interconnected. Neo4j enables users to store and navigate learning resources in a way that mirrors how knowledge is naturally structured — as a network of relationships.

This beginner-level workshop introduces participants to the core concepts of graph databases and demonstrates how Neo4j can be used to build a dynamic, searchable, and scalable learning material system. The focus is on accessibility and practical skills, using a suite of user-friendly tools including:

• Neo4j AuraDB – cloud-based graph database
• Neo4j Data Importer – visual tool for importing CSV files
• Neo4j Bloom – intuitive interface for graph exploration
• Arrows App – for modeling graph structures visually

Workshop Structure:

1. Introduction: What is a graph database and why use Neo4j?
2. Basic Theory: Nodes, relationships, and the Cypher query language
3. Tools Overview: Setting up and navigating with AuraDB, Importer, Bloom
4. Hands-on Exercises: Modeling a course structure, importing data, querying with Cypher

Expected Outcome:

Participants will gain a foundational understanding of graph databases and hands-on experience with Neo4j’s ecosystem. By the end of the workshop, they will be able to model basic educational data, import structured content, and explore it interactively—paving the way for advanced applications such as recommendation engines, adaptive learning paths, or curriculum planning.

Speaker: Vladimir Aleksic (FU Berlin)

7 A FAIR Data Space for Helmholtz: Training as Key

• Presentation of the HMC mission and the key role training plays in it
• Challenge of different needs of target groups: Data stewards/professionals, data producers, infrastructures
• What formats do we offer to reach the different target groups
• Challenge: scaling the training
• Opportunities for cooperation
• Vision: certificate course

Speaker: Mirl Trosch (Geomar Helmholtz Research Center)

8 Organising and conducting practical hands-on sessions in large online training courses

For smooth onboarding of new users to their HPC systems, the Jülich Supercomputing Centre (JSC) offers twice per year an "Introduction to Supercomputing at JSC – Theory & Practice" course. This course is the largest course in the JSC training curriculum in terms of number of participants with repeatedly over 100 registrations, and is held online to ensure the widest reach. A key component of the 3.5-day course is a series of interactive hands-on sessions designed to actively engage participants through practical exercises. The hands-on sessions take place over three consecutive mornings, starting on day two, and typically involve 30 to 60 participants per session. In this talk we detail the organisational structure for a smooth conduction of practical hands-on exercises in a large online setting that can be adapted by other online courses with practical components. Interaction with participants is organised through a multi-platform setup, employing a chat based communication system for structured real-time Q&A, and a video conferencing tool allowing for parallel-running topic discussions with domain experts. Two organisers are in charge of navigating the participants through the practical sessions, matching help requests with HPC domain experts who are present online to assist participants based-on-demand, and foster an interactive environment for topic discussions. In this talk we share best practices and guidelines for structuring communication and manage information flows, addressing the diverse needs of participants while keeping overview of participants progress and potential struggles that require assistance. We share various techniques we implement to increase and uphold engagement, which is correlated with an increased knowledge retention rate and depth of understanding, in an online setting breaking the screen-barrier.

Speaker: Jolanta Zjupa (JSC/FZJ)

18:40 Networking & Dinner

Tuesday 30 September

Workshops (swapping groups)

9 Workshop 1: Interactions, tools and methods in online teaching

Learning objectives:
Participants can develop methods and didactic concepts for online training. They can apply principles to motivate participants and get them to interact so that learning content is actively processed.

Contents of the workshop:
The workshop provides basic didactic knowledge for online teaching and introduces methods and principles that maintain a consistently high level of learning ability among participants. Special reference is made to the ‘flipped classroom’ method. In addition, the ‘Method Shower’ section of the workshop offers the opportunity to collaboratively develop your own methods together with the other participants and interactively try out tools.

Speaker: Dr Patricia Wohner (Hanke Teachertraining)

10 Workshop 2: Effective use of Large Language Models (e.g. ChatGPT) in teaching

In this hands-on workshop, participants
* learn basic techniques in prompt engineering
* get inspiration on how to use Large Language Model tools in teaching or to prep for courses.

To take part, particiants must have/create a (free) ChatGPT-Account. To make working with ChatGPT more secure, you will receive a Team license from HIDA for the duration of the workshop.

Speaker: Julian Kortendieck (HIDA)

13:00 Lunch

Parallel workshops & talks

11 Data Literacy: How to educate and train the next generation in chemistry and beyond

As we navigate an increasingly data-driven world, advanced skills in using digital tools are mandatory in day-to-day life. This digitalization has also found its way into chemical research, where more and more electronic data are being generated. Handling and management of the collected data has become a considerable part of every researcher’s work. Scientists need to be trained in these topics and concepts to apply them successfully in their research processes. To have the ability to effectively manage and utilize data, education in data literacy should start from the very beginning. However, research data topics are largely missing in chemistry curricular.

The chemistry consortium in the national research data infrastructure in Germany NFDI4Chem tackles these challenges by providing several teaching and training courses and materials. As we believe that young chemists and students are key to the cultural change, we are increasingly paying attention to education. Some lecturers, who did not want to waste time during the updating process of curricular (re-accreditation), have already introduced research data management topics in their courses. These leading examples, from institutions such as RWTH Aachen or RPTU Kaiserslautern-Landau, serve as best practice for others that are planning to integrate RDM teaching into their study programs.

The fact that data literacy education should start from the very beginning is not limited to chemistry, it applies equally to any other field. Members of several NFDI Consortia assembled to establish a platform for data literacy following the paradigm: “Data literacy for all from the very beginning!”. The formed initiative DALIA (DAta LIteracy Alliance) is developing this platform as a user-centered knowledge graph to support establishing a data culture in the academic environment. On the DALIA platform, teaching materials can be openly shared for re-use to empower others to adapt these for their individual institutions with minimal effort.

Speaker: Dr Jochen Ortmeyer (RWTH Aachen University)

12 Simplifying Cloud-Based Workshops with SimpleVM

SimpleVM is a self-service platform within the OpenStack-based de.NBI Cloud, designed to simplify access to computational resources through its customizable AAI integration. It serves different computational use-cases, including basic data processing, GPU-accelerated machine learning, and cluster computing, all secured by an intrusion detection and prevention system. By providing pre-configured Virtual Research Environments (VREs) accessible via a web browser or SSH, SimpleVM enables researchers to focus on their scientific work without requiring extensive cloud computing expertise. Our VREs come in three categories: Integrated Development Environments (IDEs) like VSCode, Data Science notebooks like RStudio or Jupyterlab, Remote Desktops, and ready-to-run research environments, e.g. for metagenomics.

One of the key features of SimpleVM is its Workshop mode, which enables easy setup and management of virtual machines (VMs) and volumes for workshop attendees. This mode is particularly well-suited for large online workshops, as demonstrated by previous SimpleVM Workshop based events with up to 50 participants. Additionally, the workshop mode is being used in a test installation at Bielefeld University as part of a data science infrastructure project.

In this presentation, we will demonstrate how SimpleVM's workshop mode can be used to streamline cloud-based training. We will showcase several key features that simplify workshop organization. Specifically, we'll demonstrate how to customize VMs from vanilla base images or our pre-configured templates with VREs. You'll also learn how to use the snapshot feature to create a template for attendee VMs. Additionally, we'll show you how to start and assign pre-configured VMs to attendees with just a few clicks. Finally, we'll cover how to provide each attendee with their individual access instructions. Participants can access pre-configured VMs with VRE installed on them easily via their browser. A live demo will illustrate the simplicity and effectiveness of SimpleVMs Workshop mode.

Speaker: Viktor Rudko (IBG-5)

13 Don't brick your research! Bluebrixx© it! Teaching usable documentation with Bluebrixx©

Reproducibility, Open Science and Documentation. The ideas are everywhere in science! In textbooks and tutorials we are repeatedly told to describe what we did so that somebody else can do it too. We try, and yet whenever anyone tries to reproduce our research, or we try to reproduce someone else’s, we find that it’s not really possible unless they contact us for details about our work, or we talk to them about their work. So then, what does it mean exactly to document research so that it is open and reproducible in such a way that that somebody else can do it without having to contact us? This workshop will use Bluebrixx (our local German flavor of the building brick toys!) to guide researchers in understanding the principles of documenting research for human users. We will outline the basic concepts of how human minds understand sequential information. We will also describe the cognitive tools we can use to make sure that information is complete and usable.

Speaker: Ms Monica Gonzalez-Marquez (Quala Lab)

14 Make a Wish: Developing FutuRSI's teaching portfolio to boost RSE education across Germany

FutuRSI is a recently started project with the aim to establish a national research software institution facilitating collaboration between research institutions and domains, coordinating services and offerings, and empowering and connecting people developing research software across Germany. The portfolio shall also include offers related to teaching - aside of courses run by FutuRSI partners, a database of workshop offers by various organisations, of freely usable teaching materials, a place to connect carpentries instructors in Germany, and more are thinkable. In this session we want to shortly introduce FutuRSI before discussing potential formats and services related to teaching. Input and feedback from the community will help to identify gaps FutuRSI could close to support educators and develop a portfolio that truly boosts the community.
From within Helmholtz, HZDR and FZJ contribute their experiences from HIFIS and JuRSE to the FutuRSI project. Further project partners are the GI, deRSE, SUB Göttingen, and Zedif in Jena. Funding comes from the Klaus Tschira Stiftung.

Speaker: Katja Linnemann (HZDR (FWCC, HIFIS))

15 Upskilling Public Sector Environmental Professionals: AI-Powered Training for Operational Use of Earth Observation Across Diverse Sectors

The increasing availability of Earth Observation (EO) data presents immense opportunities to modernize monitoring and reporting activities across public and government agencies in the environmental sector. Despite this potential, integration of EO into the workflows of municipal, regional and national administrations remains limited—constrained by a lack of time, varying technical proficiency, and fragmented access to suitable training materials. The EO4Nature project tackles this challenge by developing an intuitive online platform that tailored to environmental professionals working in public institutions. It leverages a suite of 18 interoperable EO products, enabling users to analyze environmental conditions in a sector-specific and task-oriented manner.This platform offers tailored applications and is supported by an innovative schooling concept.
The project aims to develop a flexible and personalized schooling approach that uses large language models (LLMs) and AI-driven interfaces to deliver interactive microlearning. Rather than relying on static modules, training is dynamically adapted to user expertise and specific use cases, supporting professionals in fulfilling their monitoring and reporting obligations with confidence.
Through co-design processes—including collaborative workshops, targeted surveys, and iterative testing—we ensure the training materials align with real-world workflows and institutional needs. This approach not only accelerates the uptake of EO tools but also empowers users to combine and customize services for integrated environmental insights.
EO4Nature demonstrates how modern AI-powered education can bridge the skills gap and mainstream the operational use of EO in the public environmental sector.

Speaker: Alison Beamish (GFZ Helmholtz Centre for Geoscience)

16 Building FAIR data literacy: using building-brick toys to teach the FAIR data principles

The FAIR data principles [1] are fifteen high-level guidelines aimed at improving the findability, accessibility, interoperability, and reusability of research data. Although the FAIR principles do not mandate specific technologies, they encourage the use of shared, accessible, and broadly applicable methods for representing and managing data. Research communities must have a fundamental understanding of these underlying concepts to design FAIR solutions that fit their particular disciplinary needs. Teaching these principles can be challenging, especially when engaging interdisciplinary audiences with varying technical expertise levels.
In this talk, I will present an educational approach developed at FAIRmat [2] that uses the concept of connected building bricks (e.g., LEGO®) to simplify and visualize key FAIR concepts. Each data element is represented as a brick, and key characteristics, such as detailed descriptions, shared vocabularies, standard formats, and persistent identifiers, form the connections between them. This analogy-based method has been successfully employed in FAIRmat training events [3, 4], where it increased participant engagement, reduced conceptual barriers, and facilitated a deeper understanding of FAIR principles.

References:
1. M. Wilkinson, et al., "The FAIR Guiding Principles for Scientific Data Management and Stewardship." Sci Data 3, 160018 (2016). https://doi.org/10.1038/sdata.2016.18.
2. C. Draxl et al., "FAIR Data Enabling New Horizons for Materials Research," Nature 604, 635 (2022). https://www.nature.com/articles/s41586-022-04501-x.
3. L. Rotheray et al., "FAIRmat Tutorial 11: Research Data Management, From Fundamentals to Implementation." November 30, 2023. Accessed: September 3, 2024. [Online video]. Available at: https://zenodo.org/doi/10.5281/zenodo.12760514.
4. A. E. Mansour, (2023, November 23). "Introduction to the FAIR Data Principles." Zenodo. https://doi.org/10.5281/zenodo.10200

Speaker: Ahmed Mansour (Physics Department and CSMB Adlershof, Humboldt-Universität zu Berlin, Germany)

16:00 Coffe break

Talks

17 Data Quality in the Age of AI: The Multi-Pillar Teaching Approach of the Competence Center Data Quality in the Social Sciences (KODAQS)

Data quality competencies are essential in most fields of research. Teaching skills on how to identify, foresee, diagnose, fix, and frame potential problems related to the quality of data used by social scientists is important to ensure the correctness of the conclusions drawn from the data. Especially with the rise of machine learning and large language model-based analysis approaches, high data quality is of importance - even the most advanced model can only produce outputs as meaningful as the data that it receives as input.

The KODAQS Academy, a partnership between GESIS - Leibniz Institute for the Social Sciences, the University of Mannheim, and the LMU Munich, addresses teaching data quality skills through a three-component approach: Blended learning, hands-on experience, and network building. With this submission, we present our teaching approach and report on first insights that we have gained from the completion of our first cohort throughout 2025. We identify possible avenues on how teaching can take into account recent developments in artificial intelligence (AI), regarding both teaching with and about artificial intelligence tools: Teaching with AI regards AI as assistance in our teaching approaches, enhancing (or replacing?) humans in their teaching and administrative work. In contrast, through teaching about AI, we educate researchers on how to leverage AI tools and methodologies to facilitate their research process. Especially in teaching about AI, the critical evaluation of data quality is important, as AI makes mistakes such as hallucinations or wrong conclusions, or can be biased and report one-sided representations.

After our talk and during the conference, we are especially interested in discussing and hearing the perspectives of other attendees on the two presented facets of AI in teaching: (1) Which aspects of AI are relevant when teaching data quality, and (2) How can we meaningfully integrate AI tools into our teaching.

Speaker: Florian Bemmann (LMU Munich)

18 Interactive Learning Environments: Revamping HPC Onboarding for a Seamless User Experience

The complexity of HPC environments often creates a barrier for new users, leading to inefficient utilization of resources and slower onboarding. Traditional onboarding methods rely heavily on command-line interfaces or dense documentation, which can overwhelm and discourage learners before they even submit their first job.

To bridge this gap between traditonal HPC environments and modern learning paradigms, we are leveraging Open OnDemand — an interactive web-based interface for HPC clusters. By providing users with a convinient, practical interface, the learning gets much more enganing and effortless. Instead of handling users a manual, this allows us to provide hands-on tutorials that walk learners through fundamental HPC user operations.

This platform allows learners to build competence by engaging directly with the cluster, at their own pace, and without the usual setup overhead. The result is a more intuitive and accessible entry point to HPC. By removing the initial friction and offering a guided learning experience directly on the cluster, we are able to promote best-practices and sustainable use of compute resources.

In this talk, I will provide an outlook of our learning platform and how we are transforming HPC training at HZDR — making it more inclusive, streamlined, and user-focused.

Speaker: Varun Sudharshnam (HZDR)

19 Bar Camp Session

No Pre-set Agenda
This session follows no rigid schedule—we go with the flow. You and your peers propose the topics, and together, you decide which ones to explore.

Diverse Voices Welcome
Everyone has a voice here. Share your passions, dive into dynamic discussions, and learn from the experiences of your peers.

Bring a Topic If You Want To...
* Ask a question you’re seeking answers to
* Explore a challenge you’re currently facing
* Share an innovative solution you’ve discovered
* Brainstorm ideas with others in a collaborative space

BarCamp topics might center on teaching practices, navigating the balance between everyday work and teaching commitments, experimenting with different instructional settings or tools—and much more. The possibilities are as open as the session itself.

18:45 Networking & Dinner

Wednesday 1 October

Workshops (swapping groups)

20 Workshop 1: Interactions, tools and methods in online teaching

Learning objectives:
Participants can develop methods and didactic concepts for online training. They can apply principles to motivate participants and get them to interact so that learning content is actively processed.

Contents of the workshop:
The workshop provides basic didactic knowledge for online teaching and introduces methods and principles that maintain a consistently high level of learning ability among participants. Special reference is made to the ‘flipped classroom’ method. In addition, the ‘Method Shower’ section of the workshop offers the opportunity to collaboratively develop your own methods together with the other participants and interactively try out tools.

Speaker: Dr Patricia Wohner (Hanke Teacher Training)

21 Workshop 2: Effective use of Large Language Models (e.g. ChatGPT) in teaching

In this hands-on workshop, participants
* learn basic techniques in prompt engineering
* get inspiration on how to use Large Language Model tools in teaching or to prep for courses.

To take part, particiants must have/create a (free) ChatGPT-Account. To make working with ChatGPT more secure, you will receive a Team license from HIDA for the duration of the workshop.

Speaker: Julian Kortendieck (HIDA)

22 Closing Event

Thank you for joining the TEACH!

• Summary of the conference
• Feedback

13:15 Lunch & Departure