Nick Garabedian

Dr. Nick Garabedian

orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 June 2022 in Scientific Data
Verified
Journal Article
Open access
PDF file
NGNick Garabedian
Nick Garabedian
PSPaul J. Schreiber
Paul J. Schreiber
NBNico Brandt
Nico Brandt
PZPhilipp Zschumme
Philipp Zschumme
IBInes L. Blatter
Ines L. Blatter
CGChristian Greiner
Christian Greiner
Solutions for the generation of FAIR (Findable, Accessible, Interoperable, and Reusable) data and metadata in experimental tribology are currently lacking. Nonetheless, FAIR data production is a promising path for implementing scalable data science techniques in tribology, which can lead to a deeper understanding of the phenomena that govern friction and wear. Missing community-wide data standards, and the reliance on custom workflows and equipment are some of the main challenges when it comes to adopting FAIR data practices. This paper, first, outlines a sample framework for scalable generation of FAIR data, and second, delivers a showcase FAIR data package for a pin-on-disk tribological experiment. The resulting curated data, consisting of 2,008 key-value pairs and 1,696 logical axioms, is the result of (1) the close collaboration with developers of a virtual research environment, (2) crowd-sourced controlled vocabulary, (3) ontology building, and (4) numerous – seemingly – small-scale digital tools. Thereby, this paper demonstrates a collection of scalable non-intrusive techniques that extend the life, reliability, and reusability of experimental tribological data beyond typical publication practices.

18 April 2019 in ACS Applied Materials & Interfaces
Verified
Journal Article
Closed access
KDKevin T. Dicker
Kevin T. Dicker
AMAxel C. Moore
Axel C. Moore
NGNick Garabedian
Nick Garabedian
HZHan Zhang
Han Zhang
SSSamuel L. Scinto
Samuel L. Scinto
XJXinqiao Jia
Xinqiao Jia

11 October 2019 in ACS Applied Materials & Interfaces
Verified
Journal Article
Closed access
BBB. P. Borovsky
B. P. Borovsky
NGNick Garabedian
Nick Garabedian
GMG. R. McAndrews
G. R. McAndrews
RWRaymond Wieser
Raymond Wieser
DBDavid L. Burris
David L. Burris

25 January 2022 in Data
Verified
Journal Article
Open access
PDF file
NBNico Brandt
Nico Brandt
NGNick Garabedian
Nick Garabedian
ESEphraim Schoof
Ephraim Schoof
PSPaul J. Schreiber
Paul J. Schreiber
PZPhilipp Zschumme
Philipp Zschumme
MSMichael Selzer
Michael Selzer
The ever-increasing amount of data generated from experiments and simulations in engineering sciences is relying more and more on data science applications to generate new knowledge. Comprehensive metadata descriptions and a suitable research data infrastructure are essential prerequisites for these tasks. Experimental tribology, in particular, presents some unique challenges in this regard due to the interdisciplinary nature of the field and the lack of existing standards. In this work, we demonstrate the versatility of the open source research data infrastructure Kadi4Mat by managing and producing FAIR tribological data. As a showcase example, a tribological experiment is conducted by an experimental group with a focus on comprehensiveness. The result is a FAIR data package containing all produced data as well as machine- and user-readable metadata. The close collaboration between tribologists and software developers shows a practical bottom-up approach and how such infrastructures are an essential part of our FAIR digital future.

12 October 2022 in Research Ideas and Outcomes
Verified
Journal Article
Open access
PDF file
IBIlia Bagov
Ilia Bagov
CGChristian Greiner
Christian Greiner
NGNick Garabedian
Nick Garabedian
Data's role in a variety of technical and research areas is undeniably growing. This can be seen, for example, in the increased investments in the development of data-intensive analytical methods such as artificial intelligence (Zhang 2022), as well as in the rising rate of data generation which is expected to continue into the near future (Rydning and Shirer 2021). Academic research is one of the areas, where data is the lifeblood of generating hypotheses, creating new knowledge, and reporting results. Unlike proprietary industry data, academic research data is often subjected to stricter requirements regarding transparency, and accessibility. This is in part due to the public funding which many research institutions receive. One way to fulfil these requirements is by observing the FAIR (Findability, Accessibility, Interoperability, Reusability) principles for scientific data (Wilkinson et al. 2016). These introduce a variety of benefits, such as increased research reproducibility, a more transparent use of public funding, and environmental sustainability. A way of implementing the FAIR principles in practice is with the help of FAIR Digital Objects (FDOs) (European Commission: Directorate-General for Research and Innovation 2018). A FDO consists of data, an accompanying Persistent Identifier (PID), and rich metadata which describes the context of the data. Additionally, the data format contained in an FDO should be widely used, and ideally open. Our presentation is focused on the third of FDO's components mentioned previously – metadata. It outlines the concept for a framework which enables the collaborative definition of metadata fields which can be used to annotate FDO-encapsulated data for a given domain of research. The first component of the presented framework is a controlled vocabulary of the domain related to the data which needs to be annotated. A controlled vocabulary is a collective that denotes a controlled list of terms, their definitions, and the relations between them. In the framework presented in this contribution, the terms correspond to the metadata fields used in the data annotation process. Formally, the type of controlled vocabularies used in the framework is a thesaurus (National Information Standards Organization 2010). Thesauri consist not only of the elements mentioned previously, but also allow for the inclusion of synonyms for every defined term. This eliminates the ambiguity which can occur when using terms with similar definitions. Additionally, thesauri specify simple hierarchical relations between the terms in the vocabulary, which can provide an explicit structure to the set of defined metadata fields. The most important feature of our framework, however, is that the controlled vocabularies can be developed in a collaborative fashion by the domain experts of a given research field. Specifically, people are able to propose term definitions and edits, as well as cast votes on the appropriateness of terms which have already been proposed. Despite their advantages, one limit of thesauri is their lacking capability of relating metadata fields to each other in a more semantically rich fashion. This motivated the use of the second component of the framework, namely ontologies. An ontology can be defined as “a specification of a conceptualization” (Gruber 1995). More precisely, it is a data structure which represents entities in a given domain, as well as various relations between them. After a set of metadata fields has been defined within a controlled vocabulary, that vocabulary can be transformed into an ontology which contains additional relations between the fields. These can extend beyond the hierarchical structure of a thesaurus and can contain domain-specific information about the metadata fields. For example, one such relation can denote the data type of the value which a given field must take. Furthermore, ontologies can be used to link not only metadata, but also data, as well as individual FDOs themselves. This can contribute to the Reusability aspect of FAIR Data Objects. For example, an FDO generated by a research group in a given domain can be linked to an existing domain ontology. Afterwards, the FDO can be reused more easily by researchers from the same scientific field, because the ontology will have already specified the FDO's relation to the subject area. Additionally, cross-domain ontologies can be combined with each other which can increase the reusability of FDOs beyond their domain boundaries. The components described above are being implemented in the form of multiple software tools related to the framework. The first one, a controlled vocabulary editor written as a Python-based web application called VocPopuli, is the entry point for domain experts who want to develop a metadata vocabulary for their field of research or lab. The software, whose first version is already being tested internally, enables the collaborative definition, and editing of metadata terms. Additionally, it annotates each term, as well as the entire vocabulary, with the help of the PROV Data Model (PROV-DM) (Moreau and Missier 2013) - a schema used to describe the provenance of a given object. Finally, it assigns a PID to each term in the vocabulary, as well as the vocabulary itself. It is worth noting that the generated vocabularies themselves can be seen through the prism of FDOs: they contain data (the defined terms) which is annotated with metadata (e.g., the terms' authors) and provided with a PID. The second software solution will facilitate the transformation of the vocabularies developed with the help of VocPopuli into ontologies. It will handle two distinct use cases – the from-scratch conversion of vocabularies into ontologies, and the augmentation of existing ontologies with the terms from a given thesaurus. As is the case with VocPopuli, the second tool is being developed in the Python programming language. The software solutions will be finally tested by two semi-overlapping groups of users from materials science. On the one hand, domain experts will input, edit, and discuss vocabulary terms in their area of interest, and thus create vocabularies. On the other hand, vocabulary and ontology administrators will oversee the vocabulary creation, and ontology transformation processes in a semi-automatic fashion. After development is complete, the tools will be used in the creation of controlled vocabularies for various experimental procedures, as well as their transformation and/or integration into semantically richer ontologies. This will augment our already published work in the area (Garabedian et al. 2022) and will thereby test the integration of the new framework with already existing resources. The new vocabularies will describe processes in multiple domains, such as materials science, tribology, and metalworking. Afterwards, the developed thesauri will be used in the creation of metadata templates which can be used to annotate experimental data generated in the procedures mentioned above.

26 May 2026
Verified
Preprint
Open access
PDF file
NGNick Garabedian
Nick Garabedian
WTWilfred T. Tysoe
Wilfred T. Tysoe
The scientific research community stands at a critical juncture in its approach to data management. While we have successfully transitioned from manual to computer-controlled experiments across diverse fields (from tribology to materials science, from chemistry to bioengineering), our data practices have not kept pace with the exponential growth in data generation and the emergence of artificial intelligence. This paper explores why investing time in understanding modern data architectures is essential, not just as another publication requirement, but as the foundation for accelerated collaborative discovery within our organizations and across the community. We examine the current landscape through recent survey data, explain the fundamental concepts behind FAIR (Findable, Accessible, Interoperable, and Reusable) data principles with the technical depth needed to assess true FAIR-ness, and present FAIR digital twins of scientific work as one concrete implementation that can transform isolated data into interconnected knowledge networks. Just as computer-controlled instruments revolutionized our experimental capabilities, modern data architectures promise to be the next enabling technology that amplifies our collective scientific output. While we use tribology as our primary field of interest, the principles apply equally to any experimental science dealing with multi-dimensional, context-dependent data.