Font Size: 

Manufacturer specifications for 32-channel rotary LiDARs often omit critical application-level point-cloud performance realized after system integration. A matched empirical benchmark of the Ouster OS1-32 and RoboSense Helios 32 mounted on a Clearpath Husky A200 was conducted across four standardized static and dynamic scenarios using identical processing pipelines. While both sensors demonstrated comparable static planar residuals (5.7–7.2 mm) and near-horizon vertical beam spacing (differing by 0.04°), the Helios’s 1.8X higher horizontal column density yielded smaller dimensional errors and a 1.9X precision advantage, achieving complete target-sphere detections at 9 m where the OS1-32 produced none (with neither sensor detecting targets at 10 m). Under platform motion, orienting the Helios scan boundary toward the target degraded detection and repeatability despite unaffected static baselines. Furthermore, passing vehicles (10–30 km/h) exhibited apparent length variations of 69–190 mm (Helios) and 86–217 mm (OS1-32) across opposing travel directions, demonstrating that conventional ego-motion deskewing fails to correct independent-target kinematic distortion.