Aluminothermic welding uses an exothermic chemical reaction between aluminium and iron oxide to cast molten steel into a mould around the rail joint, and it is typically used for in-track repairs where the rail cannot be moved. Flash butt welding instead clamps the two rail ends and passes a heavy electrical current through the joint to heat and then forge them together under pressure, and it is more common in rail-manufacturing plants or mobile welding trains because it produces a more uniform weld. Patent activity in this dataset covers both routes, with heavy overlap in heat-affected-zone control, preheating parameters and post-weld inspection claims regardless of which welding method is used. Choosing between them in practice depends on whether the work happens in a fixed plant or in the field.

The ranked leaders in this dataset are dominated by integrated steel producers and welding equipment manufacturers, with the top-ranked assignee holding 38 of the 314 records in scope. The top 5 companies combined account for 120 records, or 38.2% of the field, before the ranking flattens into a long tail across 94 companies total. This concentration pattern is typical of metallurgy-heavy fields: a handful of large steelmakers hold foundational alloy and heat-treatment claims, while smaller and more specialised firms file around narrower applications.

Filings rose from 4 in 2017 to a peak of 12 in 2021, then eased to 9 by 2024, a 25% decline over that three-year window. That said, patent publication typically lags the actual filing date by around 18 months, so figures for 2025 and 2026 in any dataset are still incomplete and understate real activity. The honest read is a plateau off a 2021 high rather than a confirmed downward trend, and the picture should be revisited once later years finish publishing.

US12539554B2, assigned to POSCO CO., LTD and published in February 2026, claims a flash butt welding member and method for a wheel-rim weld part, specifying a weld hardness of 220-270 Hv and geometric relationships between the heat-affected zone spacing, base metal thickness and upset length. It is aimed at wheel-rim welding rather than running-rail joints specifically, so its direct blocking effect is narrowest for parties working on rail vehicle wheel manufacture using flash butt methods with matching hardness and geometry ranges. Anyone designing a competing wheel-rim weld process should check whether their hardness range and HAZ-to-thickness ratio fall inside these claimed bounds before finalising a process spec.

The clearest gap sits between the dense metallurgy and welding-process core (B23K, C21D and C22C together cover the large majority of records) and the comparatively thin railway-system classes: E01B track & permanent way reaches only 8.6% of the 314 records, and vehicle wheels (B60B) just 2.2%. That gap suggests under-claimed ground in automated weld defect classification, in-track geometry alignment sensing, and integration of weld quality data with rail maintenance scheduling. A first claim in these areas would face a thinner prior-art field than one written around alloy composition or basic weld process steps, where the top assignees already hold dense coverage.