Resources

Researchers rely on a variety of systems and tools when it comes to administering their research data. Processes involving research data management include proposal submission, data management planning, simulation campaigns, documentation during the experiment, and the creation and submission of journal and data publications. HELIPORT is a data management solution that aims at making all steps of the research experiment’s life cycle discoverable, accessible, interoperable and reusable according to the FAIR principles. This is done by linking to and interfacing with established tools and solutions, and exchanging metadata between systems involved in a project. The metadata are presented to the researchers through a web interface, but they are also accessible to computational agents via API and machine-readable landing pages. In this presentation, we will introduce the metadata project HELIPORT and what provided the impulse for the project, discuss the documentation of a real experiment in HELIPORT, and outline current developments and challenges.

HELIPORT is a data management guidance system that aims at making the components and steps of the entire research experiment's life cycle findable, accessible, interoperable and reusable according to the FAIR principles. It integrates documentation, computational workflows, data sets, the final publication of the research results, and many more resources. This is achieved by gathering metadata from established tools and platforms and passing along relevant information to the next step in the experiment's life cycle. HELIPORT's high-level overview of the project allows researchers to keep all aspects of their experiment in mind.

A particularly interesting use case are machine learning projects. They are often prototypical in nature and driven by iterative development, so reproducibility and tranparency are a great concern. It is essential to keep track of the relationship between input data, choices in model parameters, the code version in use, and performance measures and generated outputs at all times. This requires a data management platform that automatically records the changes made and their effects. Existing MLOps tools (such as Weights and Biases, MLFlow) live entirely in the ML domain and start their workflow with the assumption that data is available. HELIPORT, on the other hand, takes care of the data lifecycle as well. Our envisioned platform interoperates with the domain specific tools already used by the scientists, and is able to extract relevant metadata (e.g. provenance). It can also make persistent any additional information such as papers the work was based on, documentation of software components, workflows, or failure cases. Moreover, it should be possible to publish these metadata in machine-readable formats.

The challenge arising from these aspects consists in integrating ML workflows into HELIPORT in such a way that they work on the provided data and metadata. The goal is also to enable the comprehensible development of ML models alongside the experiment documented in HELIPORT. This allows different teams (e.g. experimentalists and AI specialists) to work together on the same project in a seamless manner, and help generate FAIRer outcomes. In the long term we hope to aide in establishing digital twins of facilities, and making their maintenance a part of the data management proces.

Researchers at the Helmholtz-Zentrum Dresden-Rossendorf rely on a variety of systems and tools when it comes to administer their research data. Processes involving research data management include the project planning phase (proposal submission to the beamtime proposal management system, the creation of data management plans and data policies), the documentation during the experiment or simulation campaign (electronic laboratory notebooks, wiki pages), backup- and archival systems and the final journal and data publications (collaborative authoring tools, meta-data catalogs, software and data repositories, publication systems). In addition, modern research projects are often required to interact with a variety of software stacks and workflow management systems to allow reproducibility on the underlying IT infrastructure. The "HELmholtz ScIentific Project WORkflow PlaTform" (HELIPORT), which is currently developed by researchers at HZDR and their collaborators, tries to facilitate the management of research data and metadata by providing an overarching guidance system which combines all the information by interfacing the underlying processes and even includes a workflow engine which can be used to automate processes like data analysis or data retrieval.

Digital infrastructures have become indispensable in the field of modern research and science. These technological frameworks play a crucial role for the entire research cycle, supporting literature searches, aiding in data collection and analysis, facilitating the creation and publication of scholarly works, and ensuring the thorough documentation and long-term storage of research findings. Additionally, these infrastructures serve as a vital means for networking and communication among peers, creating the essential foundation of an open and transparent science and research ecosystem.
In this lecture, the entire digital research landscape at the HZDR will be presented and illustrated using a representative experiment.

In this presentation, Heliport is outlined with a focus on the research software product Heliport itself. The project was initially funded by the HElmholtz Metadata Collaboration (HMC).

When dealing with research data management, researchers at Helmholtz- Zentrum Dresden – Rossendorf (HZDR) face a variety of systems and tools. These range from the project planning phase (proposal management, data management plans and policies), over documentation during the experiment or simulation campaign, to the publication (collaborative authoring tools, metadata catalogs, publication systems, data repositories). In addition, modern research projects usually are required to interact with a variety of software stacks and workflow management systems to allow comprehensi- ble and FAIR science on the underlying IT infrastructure (HPC, data storage, network file systems, archival). This article first demonstrates the data management systems and services provided at HZDR, followed by an overview of a self-developed guidance system. It is concluded by a real-world example.

The HELIPORT poster provides a short overview on the project and introduces the usage of Handles, the workflow integration, the top-level project plan and the project metadata schema.

The presentation outlines the POLARIS experiment at Helmholtz Institute Jena, including experimental chain, setup and first ideas regarding the description of the POLARIS experiment with HELIPORT.

The presentation outlines the HELIPORT project. The HELIPORT project aims at developing a platform which accommodates the complete life cycle of a scientific project and links all corresponding programs, systems and workflows to create a more FAIR and comprehensible project description.

Modern scientific collaborations and projects (MSCPs) employ various processing stages, starting with the proposal submission, continuing with data acquisition and concluding with final publications. The realization of such MSCPs poses a huge challenge due to (1) the complexity and diversity of the tools, (2) the heterogeneity of various involved computing and experimental platforms, (3) flexibility of analysis targets towards data acquisition and (4) data throughput. Another challenge for MSCPs is to provide additional metadata according to the FAIR principles for all processing stages for internal and external use. Consequently, the demand for a system, that assists the scientist in all project stages and archives all processes on the basis of metadata standards like DataCite to make really everything transparent, understandable and citable, has risen considerably. The aim of this project is the development of the HELmholtz ScIentific Project WORkflow PlaTform (HELIPORT), which ensures data provenance by accommodating the complete life cycle of a scientific project and linking all employed programs and systems. The modular structure of HELIPORT enables the deployment of the core applications to different Helmholtz centers (HZs) and can be adapted to center-specific needs simply by adding or replacing individual components. HELIPORT is based on modern web technologies and can be used on different platforms.