Which railway track maintenance machine measures geometry?

The three machine formats — and what each one is for

The mistake buyers make is treating these as interchangeable. A geometry car tells you which kilometres are getting worse; it does not give you the per-sleeper adjustment plan a tamper needs. A static system certifies absolute position, but it will not cover a 20 km possession in a night.

What the numbers actually are, so you can compare machines

Class-1 relative trolleys are quoted at around ±0.50 mm on gauge and superelevation — normally enough to plan tamping work. Where the record has to be accepted as an engineering handover, you step up to class-0 relative or to absolute measurement against a control network. Measurement efficiency separates the sensor principles as well: about 1.5–3 km/h for inertial measurement, 4–5 km/h for GNSS-augmented work, 1–2 km/h for the dual-sensor short-wave survey, and 4–6 km/h for a class-relative trolley.

How to choose — three questions that decide it

1. What has to be certified? If the output is a handover or acceptance record, you need class-0 relative or absolute referencing. If the output is a work order, class-1 relative measurement is enough and gets you round the track far faster.
2. What is the environment? GNSS-augmented trolleys want open sky and daylight; in tunnels, deep cuttings or under dense structures, a self-contained inertial trolley measures without GNSS or track markers.
3. Short wave or long wave? Short-wave irregularity — the kind that drives noise and rolling contact fatigue — needs the slow contact pass at 1–2 km/h. Long-wave alignment needs the 300 m chord or a satellite reference. One machine rarely does both ends well.

On the hardware side, the trolley itself is deliberately light: a class-relative Molaton-brand RCI GJY-series unit measures 1705 × 850 × 935 mm, weighs ≤ 50 kg and takes about 5 minutes to set up on track. Its output is already a work document — non-compliance reports, curve inspection reports, track inspection reports and a static Track Quality Index report for acceptance.

Case and data — what the measurement pass is worth

On the Taiyuan railway administration's Taizhong down line, a measured fine-tamping campaign across nine work sections totalling 15.28 km moved the average static TQI from 7.19 to 4.52 (a reduction of 2.67) and the average dynamic TQI from 7.64 to 4.67 (a reduction of 2.97). None of that is possible without the measurement pass: the trolley data is processed into a tamping plan at 2, 2.5 and 5 m intervals, exported as VER/TGCS files and imported directly into the tamping machine's computer. After tamping, the same track is re-measured and the static TQI report becomes the acceptance document.

The same logic closes on the grinding side, which is why measurement belongs in a rail grinding wheel programme: after grinding, acceptance is judged by measured values — surface roughness Ra ≤ 10 μm, 0.3 mm or less under a 1 m straightedge, no blue burning, grinding depth held to ≤ 0.5 mm and the finished profile blended out at better than 1‰. Verification on a Loram DM01 machine in 2026 returned no blue burning and roughness of Ra 1.05–9.0 μm, mostly 2–4 μm, on wheels made by RailwayCare — which has produced these wheels since 2004, the first dedicated manufacturer in China, to the industry standard JB/T 11431 the company drafted.