University of Thessaly applied-ML researcher bridging predictive maintenance and scientific machine learning
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
5
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-13) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
Unscored for now: the snapshot lacks the data needed to compute a publication-output measure of volume, venue mix and growth over time; yearly counts exist, but they do not yet support a reliable score.
Impact—
Impact cannot be scored yet: citation data in the snapshot is too thin for a measure of citation volume, field-weighted influence and h-index trajectory; h-index 5 and i10-index 5 are recorded, but the axis stays unrated.
Focus0/100
Scores 0/100 on topical concentration: the field distribution splits exactly 50/50 between predictive maintenance / applied machine learning and computational mechanics / scientific machine learning, and OpenAlex topics fan out across COVID-19 epidemiological studies (4 works), tissue engineering and regenerative medicine (4), soft robotics (3), assistive technology (3) and advanced neural network applications (2) — broad spread rather than one concentrated line.
Collaboration39/100
Moderate at 39/100: only 6 distinct co-authors appear in the snapshot, with a small recurring circle — Dimitris K. Iakovidis, Dionisis Kalogeropoulos, Georgia Sovatzidi and Georgios Triantafyllou. Both Evolve-reviewed papers position him as a middle author without corresponding status, consistent with a collaborative contributor inside a compact team rather than a wide international network.
Activity72/100
Relatively strong at 72/100: 11 works in the most recent three-year window, with yearly counts peaking at 6 in 2024 and 4 already indexed in 2026, following 3 in 2022 and 3 in 2023. Earlier coverage is sparse — 2 works in 2010 and singles in 2014, 2017, 2020 and 2021 — so the current cadence marks a clear step-up in output.
Reproducibility20/100
Low at 20/100 (reproMean across 2 papers): both Evolve-reviewed preprints carry low repro-readiness. DeepFEAv2 reports large gains over DeepFEA but its comparison is confounded and lacks ablations; the naval-maintenance paper does release public CODLAG data with 99.24% reported accuracy, yet baseline-comparison and leakage concerns remain unresolved.
Research summary
Panagiotis G. Kalozoumis is a University of Thessaly (Greece) researcher whose indexed output splits evenly (0.5/0.5) between predictive maintenance / applied machine learning and computational mechanics / scientific machine learning. Two 2026 arXiv preprints reviewed by Evolve illustrate both tracks. DeepFEAv2 extends deep-learning surrogates to transient finite element analysis on unstructured meshes using ConvLSTM and autoencoder components, reporting large gains over DeepFEA — though the review flags confounded comparisons and missing ablations. The second preprint builds an explainable condition-based maintenance framework for naval-propulsion systems, distilling a residual neural network with a fuzzy head into local fuzzy decision trees and reporting 99.24% accuracy on public CODLAG data, with baseline and data-leakage concerns noted. OpenAlex topics add breadth: COVID-19 epidemiological studies (4 works), tissue engineering and regenerative medicine (4), soft robotics (3), assistive technology (3), and advanced neural network applications (2). Activity is comparatively strong (72/100): 11 works in the most recent three years, peaking at 6 in 2024, with 4 already indexed in 2026. Collaboration is modest (39/100) with 6 distinct co-authors and recurring partners including Dimitris K. Iakovidis and Georgios Triantafyllou; both reviewed papers list him as a middle author without corresponding status. Reproducibility is currently low (20/100 across 2 papers), while the snapshot records an h-index of 5 and an i10-index of 5.
Collaboration pattern
Collaboration is compact: 6 distinct co-authors in the snapshot, with a recurring core of Dimitris K. Iakovidis, Dionisis Kalogeropoulos, Georgia Sovatzidi and Georgios Triantafyllou. In both Evolve-reviewed 2026 preprints he appears as a middle author without corresponding status, suggesting he contributes technical depth within a small, stable, University of Thessaly-centered team rather than leading large or international consortia. The collaboration radar score is 39/100.
Research keywords
Highlights
DeepFEAv2: transient FEA surrogates beyond structured meshes
Explainable condition-based maintenance for naval propulsion
A deliberately two-track machine-learning portfolio
Indexed output splits evenly (0.5/0.5) between applied ML for predictive maintenance and scientific ML for computational mechanics, with OpenAlex topics additionally spanning COVID-19 epidemiology, tissue engineering, soft robotics and assistive technology — a spread that drives the focus score of 0.
Timeline
- 2010Earliest works in the OpenAlex snapshot: 2 indexed publications.
- 2022Output picks up: 3 indexed works, up from 1 in 2021.
- 2024Most productive indexed year in the snapshot: 6 works.
- 20264 works indexed so far, including both Evolve-reviewed arXiv preprints (DeepFEAv2; explainable naval-propulsion maintenance).
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- All bibliometrics are derived from OpenAlex-indexed records only (identity resolved via ORCID 0000-0002-0208-301X).
- Output and impact axes are unscored (dimMissing); the quoted h-index of 5 and i10-index of 5 come from the snapshot and are not independently verified.
- The reproducibility score averages only the 2 Evolve-reviewed papers, both 2026 arXiv preprints rated low repro-readiness; verdicts are initial reviews.
- Yearly counts (2010–2026) reflect indexing coverage in the snapshot and may undercount earlier or non-indexed output.