A Multimodal Railway Vibration and Vision Dataset (Rail-VIVID)

Sizhe Ma, Katherine A. Flanigan, Mario Bergés, James D. Brooks · 2026

Read A Multimodal Railway Vibration and Vision Dataset (Rail-VIVID) by Sizhe Ma, Katherine A. Flanigan, Mario Bergés, James D. Brooks (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13601582.txt.

Description

Sizhe Ma and three colleagues release Rail-VIVID in 2026, a paired dataset meant to push railway track monitoring forward. Trains can inspect track as they run, using vibration sensors that feel hidden defects but cannot locate them, or cameras that show surface damage but miss what lies beneath. The two approaches grew up apart, partly because no public dataset paired both signals for direct comparison.

Rail-VIVID records synchronized vibration and video from repeated runs over one track segment with documented faults. Researchers can use it to study how each kind of fault signs itself in each signal, and to benchmark detection algorithms on the same material. It also supports building fusion methods that combine both streams for sturdier detection. The authors frame the dataset as groundwork for predictive maintenance, where catching faults early from ordinary train runs could replace some manual inspection.

Catalog record

CREATORSizhe Ma, Katherine A. Flanigan, Mario Bergés, James D. Brooks
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE41895
LICENSEhttps://creativecommons.org/licenses/by/4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/article/MED/PMC13601582

Browse the full library · About the Tower of Babel