ML robustness researcher building multi-domain benchmarks for virtual sensing under sensor failures
1
papers here
1
reviewed by Evolve
1
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-20) 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—
Not yet scored: this axis measures overall publication volume and output trajectory, but no bibliometric productivity data is available beyond the single 2026 arXiv preprint reviewed here, so no estimate is made.
Impact—
Not yet scored: this axis would capture citations, usage, and uptake of the researcher's work. No citation or influence data is available yet for the 2026 benchmark paper, so no score is reported.
Focus100/100
Scores 100: every reviewed record sits in a single field — machine learning robustness benchmarks for time-series and cyber-physical systems. The field distribution contains exactly one entry (one paper), so the concentration is total but rests on a very small base; the 2026 MuViS-C benchmark defines the whole topical footprint so far.
Collaboration39/100
Scores 39, based on 6 distinct coauthors across the available records. The network is small and tight-knit: recurring collaborators include Noah C. Puetz, Alexander Windmann, Marc Hilbert, Elena Raponi, Thomas Bäck, and Thomas Bartz-Beielstein, suggesting work within a stable research group rather than broad cross-institutional partnerships. The sole cataloged paper is first-authored without a corresponding role.
Activity—
Not yet scored: this axis tracks publication cadence and research activity over time. The catalog currently holds only one paper from 2026, which is too little to establish any temporal pattern, so the axis is left unscored.
Reproducibility100/100
Scores 100, but from a single reviewed paper (reproPapers = 1, mean 100). The 2026 MuViS-C virtual-sensing benchmark is rated high for reproducibility readiness, meaning its artifacts and setup support replication; Evolve's review still flags the single-seed experimental design as a limitation, so the perfect score should be read as one positive data point, not an established track record.
Research summary
Jens U. Brandt is an early-stage machine-learning researcher whose public bibliographic footprint currently centers on one substantial contribution: the 2026 arXiv preprint 'Reliable Virtual Sensing: A Multi-Domain Benchmark for Robustness Under Sensor Failures', on which he is first author. The paper introduces MuViS-C, a multi-domain benchmark that stress-tests virtual sensing models under sensor failures and distribution shift. Evolve's review judges it a credible benchmark showing that all evaluated virtual sensors fail under corruption, with tree-based models the most robust family, while noting that the evidence rests on single-seed experiments. Reproducibility readiness is rated high, and the reproducibility dimension scores 100 from that single reviewed paper. The profile shows extreme topical concentration (focus score 100) within machine-learning robustness benchmarks for time-series and cyber-physical systems, and a small collaboration network of six distinct coauthors (collaboration score 39), including recurring collaborators Noah C. Puetz, Alexander Windmann, Marc Hilbert, Elena Raponi, Thomas Bäck, and Thomas Bartz-Beielstein. Output, impact, and activity dimensions currently lack sufficient data and are left unscored, and no institutional affiliation or country is confirmed.
Collaboration pattern
Brandt's visible network is compact: 6 distinct coauthors (collaboration score 39), anchored by recurring collaborators Noah C. Puetz, Alexander Windmann, Marc Hilbert, Elena Raponi, Thomas Bäck, and Thomas Bartz-Beielstein. The pattern suggests a single stable research group producing benchmark-driven machine-learning work; the reviewed 2026 paper is first-authored, non-corresponding. No cross-institution or industry collaboration patterns can be inferred from the current data.
Research keywords
Highlights
MuViS-C: a multi-domain robustness benchmark for virtual sensing
No virtual sensor survives corruption — tree models hold up best
High reproducibility readiness for the MuViS-C benchmark
Timeline
- 2026Releases first-author arXiv preprint 'Reliable Virtual Sensing: A Multi-Domain Benchmark for Robustness Under Sensor Failures'.
- 2026Introduces MuViS-C, a multi-domain benchmark stress-testing virtual sensing under sensor failures and distribution shift.
- 2026Evolve review reports all tested virtual sensors fail under corruption, with tree models most robust and single-seed evidence flagged.
- 2026Reproducibility readiness rated high; the reproducibility dimension scores 100 from this single reviewed paper.
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.
- Bibliometric dimension scores are computed from OpenAlex-derived records only; the evidence base is one reviewed paper and 6 distinct coauthors.
- No ORCID, institutional affiliation, or country is confirmed; the profile is name-resolved from public bibliographic data.
- Output, impact, and activity dimensions are unscored because no citation or productivity data is available yet.
- The reproducibility score of 100 rests on a single paper (reproPapers = 1), and the reviewed evidence is single-seed.