One assignee leads the ranked field with 19 of the 57 records in scope, well ahead of the rest of the ranking of 22 companies. The top five assignees combined account for 87.7% of all records, and the top ten account for 100.0% — meaning the entire dataset is represented within the first ten ranked filers. This is a concentrated field rather than one with a long tail of small independent filers.

Filing peaked at 10 records in 2021 and fell to 1 by 2024, a 90% decline across that span. This does not necessarily mean the reliability problem is solved; it more likely reflects that the obvious mechanical fixes and lubricant formulations were already claimed by the leading filers earlier in the window. Note that 2025 and 2026 figures in any dataset like this are understated because patent publication typically lags filing by about 18 months, so a further apparent decline in those years should not be read as continued slowdown yet.

Both are documented failure modes for wind turbine gearbox bearings and gears, but they sit in different parts of the claim space in this dataset. White etching cracks relate to sub-surface material degradation often linked to electrical or lubricant-driven mechanisms, while tooth flank fracture is a structural fatigue failure tied to load spectrum and torque reversal conditions. Filings addressing either tend to cross into the C10M/C10N lubricant subclasses or the F16C shaft-and-bearing subclass depending on which mechanism they target.

The technology composition data shows F03D (wind motors generally) covering 47.4% of the 57 records, while more specific subclasses — lubricants at 17.5%, shafts and bearings at 15.8%, vibration measurement at 10.5% — are comparatively thin. A claim drafted narrowly around lubricant cleanliness thresholds, shaft misalignment sensing, or white etching crack detection methods is likely to face less crowded prior art than one framed broadly at the turbine level.

No. High filing density in a subclass like F03D shows that claim space is occupied, not that the underlying technology is the most effective or mature solution. Citation counts on individual records, such as the most-cited entries in this dataset, are a signal of influence within the searched corpus and tend to favour older filings — they are not a reliable proxy for current commercial importance. Decisions about which technical route to pursue should weigh the abstract and claims of specific records, not just how many patents exist near them.