Projects

Open Research Data Projects

Projects funded in the framework of the ORD Program

The joint ORD program of ETH Zurich, EPFL and the four research institutes of the ETH Domain has financially supported more than 60 research projects in the period 2020–2023. Funding supports researchers engaging in, or developing, ORD practices with and for their community and assists these researchers in becoming Open Research Data leaders in their field.

This page provides an overview of these projects. It highlights how researchers in the ETH Domain are currently applying ORD in exemplary ways. Some of the projects have already been completed, others are still in progress. The projects have been divided into three categories.

“Establish” projects help link existing ORD practices to a research agenda to establish them on a broader basis. They contribute to a shared and comprehensive understanding of ORD practices that can then become de facto standards.

“Explore” projects are the most extensive ventures in the program and are designed to explore and test early-stage ORD practices. The goal is to map processes of what an ORD practice might look like and develop prototypes. Through these projects, new teams form across disciplines and institutions.

“Contribute” projects help scientists integrate their research data into existing, often international, infrastructures. By standardizing the processes and making them generally accessible, the data are validated, and their potential is considerably expanded.

Filter

Category
Category
Institutions
MMS (Masonry MicroStructures database) - A 3D masonry microstructures database for advancing numerical research on irregular stone masonry structures

Category

Contribute

Institutions

EPFL

Data type

Microstructure database

Field

Materials Science

Researchers

Shah, Mati Ullah

Abstract

Stone masonry is an eco-friendly construction material, but its use has declined due to its vulnerability to earthquakes, mainly because of the poor arrangement of its microstructure. The microstructure includes the shape, size, and arrangement of stone units, which vary based on geographic, temporal, and material factors. Current building codes cannot fully account for this variability, and experimental studies are costly and impractical due to the diversity of masonry typologies. Numerical studies offer a solution, but creating realistic microstructures for modeling irregular stone masonry is complex and time-consuming. As a result, simplified microstructures are often used in simulations, which fail to capture the complexities of irregular masonry walls. To address this challenge, we have developed a 3D masonry microstructures database ready to use in numerical simulations. To enhance accessibility and usability, this project aims to create a web-based platform hosting this curated database of 3D microstructures and their geometric indices. The proposed web-based platform will also feature a tool for evaluating masonry quality using the Masonry Quality Index (MQI) from 2D images, promoting the preservation of historic structures and sustainable construction practices. Additionally, the platform will enable researchers to contribute and document new 3D microstructures, fostering collaboration and advancing numerical research on stone masonry.

Application Programming Interface for the River to Ocean Geodatabase for Education and Research

Category

Contribute

Institutions

ETH Zurich

Data type

Environnement

Field

Earth sciences

Researchers

Paradis, Sarah

Abstract

In order to advance our understanding of the carbon cycle, it is essential to evaluate the spatiotemporal variations of carbon between river and marine environments and gain insights into the pathways of carbon transfer from land to ocean. To do this, we need to work jointly with riverine and marine data, accounting for their temporal and spatial distribution. However, each of these systems have different data and metadata reporting strategies that need to be accounted for, which complicates their joint application. Efforts have been made to compile data from each of these systems into independent databases, but no attempt has yet been done to create a joint database of data of both of these systems while accounting for their different metadata. Hence, this project aims to bring together riverine and marine data into one database to easily query the data between both systems through the River to Ocean Geodatabase for Education and Research (ROGER). This database will be displayed in an interactive web-interface that queries riverine and/or marine data depending on the user’s requirements through a REST API. Harnessing the advanced geographical functions of PostgreSQL, the REST API will include functions that allow users to geospatially integrate riverine and marine data. This new database will provide a crucial step forward in the understanding of the carbon cycle along the land-ocean continuum, while ensuring that the data complies with best Open Research Data practices.

Development of standardized Respiratory Open Access Research

Category

Contribute

Institutions

EPFL

Data type

Medical data

Field

Life sciences

Researchers

Dan, Jonathan

Abstract

Chronic cough is a common condition globally. While efforts are being made to develop wearables to detect and quantify cough events automatically, such monitoring devices have not yet been incorporated into routine clinical practice due to a lack of consistency in their validation, resulting in slow progress and a lack of trust in reported results. We have identified three main reasons for this heterogeneity: 1) the clinical definition of different cough events and especially the delimitation of their beginning/end lacks standardization, 2) the data used is typically private and imbalanced with inadequate labelling as a result of the previous point, and 3) methodologies to assess the accuracy of event detection are different between research groups and often inappropriate. This proposal builds on ORD datasets, community guidelines, and standards to propose a unified framework for validating cough event detection algorithms. The main objective is the development of standards that will unify the workflow for validating respiratory event detection algorithms to ensure data adheres the principles of Findable, Accessible, Interpretable, and Reusable data. This will be distributed through a website, serving as a central hub and reference for standardizing clinical definitions and methodologies, leading to a future benchmarking platform for respiratory event detection algorithms.

Filter

Category
Category
Institutions
Expanding SRM as a unified hub for open and collaborative super-resolution microscopy

Category

Contribute

Institutions

EPFL

Data type

Super-resolution microscopy (SRM)

Field

Bioengineering

Researchers

Wayne Wen Wei Yang

Abstract

The super-resolution microscopy (SRM) hub website shares SRM resources and facilitates benchmarking competitions, primarily focused on single-molecule localization microscopy (SMLM). To revive user participation, we plan to expand the site as an inclusive platform for SRM education, dataset sharing, and software resources. We will keep content current with modern techniques and provide user-friendly access through Jupyter notebooks and instructional videos. This will empower multidisciplinary researchers. Educational presentations and workshops will increase engagement. Standardized benchmarks will aid algorithm selection. User feedback will improve resource quality and foster collaborations among creators and users of shared datasets and software on the SRM hub.

AI module for gap-filling TreeNet time series

Category

Contribute

Institutions

WSL

Data type

TreeNet dendrometer data

Field

Forest Dynamics

Researchers

Mirko Lukovic

Abstract

In a recent WSL research project (deepT - internal grant no. 202011N2099), a machine learning model was developed for gap-filling multi-channel time series data. The goal is to incorporate this model into the automated near real-time TreeNet data acquisition infrastructure. This addition will allow TreeNet dendrometer data users to fill time series gaps automatically using artificial intelligence. The existing model needs refinement with new data, programming in R (the native language of TreeNet software), integration into the pipeline, and adaptation of data to the model's input and output requirements.

Seamlessly containerised physics analysis workloads

Category

Contribute

Institutions

PSI

Data type

Particle physics

Field

Physics

Researchers

Clemens Lange

Abstract

Particle physics analyses face reusability challenges due to manual, undocumented steps by analysis teams. Typically, reusable datasets are processed with individually developed scripts and frameworks. This project addresses these issues for PSI Laboratory for Particle Physics (CMS, Mu3e, n2EDM) experiments by encoding missing information in the code repository and integrating it into the execution environment. It provides templates for automatic code testing and packages it in portable software containers. These containers are distributed on compute clusters for parallel processing. With a focus on user experience and adoption, this setup builds on PSI's existing infrastructure, aiming for large-scale deployment. It includes documentation, training, and sharing results at an international computing workshop.

Make speleothem data more FAIR: facilitating addition of SISAL to Neotoma/ LiPD

Category

Contribute

Institutions

ETH Zurich

Data type

Paleoclimate archives

Field

Spatial and Landscape Development

Researchers

Laura Endres

Abstract

The power and pace of climate research are being greatly increased by maximizing the volume and quality of openly accessible data. Vital sources of climate data are paleoclimate archives such as speleothems. Robust infrastructures are crucial for global analysis across time scales. Community-curated data resources have proven to be the most successful solution, establishing science-driven data standards and wide community support. The SpeleoFAIR project integrates SISAL with Neotoma and LiPD, strengthening data accessibility and governance. This aligns with broader goals of enhancing data sustainability and accelerating multi-proxy climate research analysis.

Global Geodetic VLBI Data Hub

Category

Contribute

Institutions

ETH Zurich

Data type

Very Long Baseline Interferometry (VLBI)

Field

Earth sciences

Researchers

Matthias Schartner

Abstract

Geodetic Very Long Baseline Interferometry (VLBI) is vital for precise Earth rotation and orientation measurements. The International VLBI Service for Geodesy and Astrometry (IVS) collaboratively collects VLBI-related files from global institutions. This project consolidates performance metrics from IVS submissions into a common database with session-, station-, and source-based tables. It offers open access through a web interface and API, enabling quick program adjustments with online plotting tools. These databases optimize observing plans by excluding underperforming elements. This enhances data accuracy and fosters VLBI community collaboration, advancing open research data initiatives in Earth sciences and geodesy, particularly for the IVS community.

OGAIS – An Open Georeferencing of Airborne Image Spectrometry

Category

Contribute

Institutions

EPFL

Data type

Hyperspectral data

Field

Geodetic Engineering

Researchers

Jan Skaloud

Abstract

This project aligns with the development of a high-resolution airborne image spectrometer (AIS) by NASA-JPL for Swiss research institutions. It aims to create tools for precise georeferencing of hyperspectral data, vital for measuring Earth system processes regionally. Outcomes include: a) Introducing an open sensor stabilization format in collaboration with ISPRS. b) Developing a tool for annotating known data points. c) Expanding EPFL's public service for sensor-motion estimation and calibration. d) Releasing reference-processed datasets through peer-reviewed publication and webinars.

emPowering RESEarch in urbaN mobility with interactive Traffic dATa visualizatION

Category

Contribute

Institutions

EPFL

Data type

pNEUMA

Field

Traffic engineering

Researchers

Emmanouil Barmpounakis

Abstract

The project's main goal is to develop a visualization platform for the pNEUMA dataset, a significant urban traffic data collection from drones. This dataset benefits global researchers, offering both macroscopic and microscopic traffic insights. Due to its vast scale, users struggle to identify relevant portions. Our platform aims to provide an open tool with maps and traffic metrics like vehicle counts, speed, and flows. An initial prototype exists, with plans to expand and improve user interactivity, data filters, and visualization for better traffic understanding. These enhancements empower users to quickly grasp traffic conditions, benefiting traffic engineering and sustainable transportation research.

Making Human Brain And Behavior Data Accessible For Modeling via Brain-Score

Category

Contribute

Institutions

EPFL

Data type

Neural and behavioral data

Field

Neuroscience

Researchers

Martin Schrimpf

Abstract

Brain-Score, a rapidly expanding platform, compiles a wide array of neural and behavioral data from non-human primate neuroscience experiments. It facilitates the evaluation of computational models of the brain's visual system. By translating machine learning models into neuroscientific hypotheses and converting neuroscience data into quantitative benchmarks, Brain-Score offers both computational and experimental advancements to the broader community synergistically. This proposal aims to expand Brain-Score to include data from human experiments, making it accessible to the cognitive neuroscience community and enhancing the applicability of computational models. The objectives are as follows:

  • Develop software for human data integration.
  • Curate the Natural Scenes Dataset (NSD) for reproducible model evaluations.
  • Include the THINGS-similarity dataset for accessible and reproducible model assessments.
Boost for the validation of seizure detection algorithms with ORD (Boost4Epilepsy)

Category

Contribute

Institutions

EPFL

Data type

Electroencephalography (EEG)

Field

Electrical and Computer Engineering

Researchers

David Atienza

Abstract

In the past decade, numerous initiatives have provided annotated scalp EEG datasets of individuals with epilepsy for global researchers. However, data format discrepancies and varying validation methods impede algorithm comparisons, rendering ranking and research progress challenging. This proposal extends ORD datasets, community standards, and guidelines to introduce a unified framework for validating seizure detection algorithms. The primary goal is to create tools and standards streamlining the algorithm validation process, aligning data with Findable, Accessible, Interpretable, and Reusable (FAIR) principles. To establish this framework, existing public annotated datasets will be curated, and their data format standardized. Software tools will be developed to convert original datasets into this standardized format. Methodologies for algorithm evaluation will also be standardized to create a benchmark for cutting-edge algorithms. All this knowledge, datasets, standards, tools, and benchmarks will be centralized on a website, expediting seizure detection algorithm development.

MAST (MAsonry Shake-Table database) - A comprehensive database and collaborative resource for advancing seismic assessment of unreinforced masonry buildings

Category

Contribute

Institutions

EPFL

Data type

Experimental campaigns

Field

Earthquake engineering

Researchers

Mathias Haindl Carvallo

Abstract

Every year, building collapses due to earthquakes cause thousands of deaths. Due to the lack of design to resist seismic forces, unreinforced masonry buildings are very prone to collapse. By developing an open-access web-based platform that will contain the first-ever consolidated database for shake-table tests on complete unreinforced masonry buildings, this project aims to contribute to addressing this critical issue. The platform will make comprehensive data from over seventy experimental campaigns conducted over the last 30 years easily accessible to the earthquake engineering community, including researchers and practitioners. As a result, by facilitating data exchange and collaboration, this platform will play a key role in advancing our understanding of the seismic behavior of unreinforced masonry structures. The platform may be used by researchers and practitioners as a valuable reference to benchmark their models, enhance predictive capabilities, and encourage improved design and retrofitting practices. Overall, the development of this platform has enormous potential to reduce seismic risk and increase the resilience of unreinforced masonry buildings, hence increasing public safety in earthquake-prone areas.

Scroll to Top

Filter

Category
Category
Institutions