Within this 49-record dataset, one assignee leads the ranking with 14 records, well ahead of the fifth-ranked filer at 6 and the tenth-ranked filer at 2. The ranking covers 25 companies in total, so it is not a top-50 or top-100 list — it is the full set the data returns. That leader’s advantage is concentrated in core impeller and volute geometry claims rather than spread evenly across every sub-area.

Specific speed is one of the search terms used to build this corpus because it is the standard way engineers characterise a centrifugal pump’s hydraulic operating point, and many impeller and volute claims are written with reference to it. Patents that specify blade angle profiles, such as the representative CA2833193C filing, typically tie the claimed geometry to a target specific-speed range rather than claiming a shape in isolation. That means freedom-to-operate checks should look at the specific-speed context a claim is written for, not just the blade shape.

The IPC composition shows F04D (non-positive-displacement pumps) covering 87.8% of the 49 records, while classes touching computational and AI-assisted design — G06F at 10.2% and G06N at 4.1% — are far thinner. That gap suggests CFD-driven optimisation and machine-learning-assisted impeller trimming are less claimed than the underlying mechanical geometry itself, making them a more open area for a first-mover claim. Radial thrust prediction and variable-speed volute matching also show limited direct representation in the most-cited records.