The compact, 20-channel solid-state LiDAR supports close-range obstacle detection as robots travel along the rails.

Benewake’s NOVA20 fully solid-state line-array LiDAR has been deployed on railway maintenance robots that automatically locate, tighten and loosen rail fasteners. Mounted at the front and rear of each robot, the sensors provide distance measurements to support collision avoidance in both directions of travel.

Detecting Obstacles in Time to Respond

Rail fastener maintenance involves repetitive work across large numbers of bolts, often within limited maintenance windows. Robots can automate bolt tightening and loosening, but they must also move between work positions without colliding with equipment or other obstacles along the track.

Rail-mounted robots cannot steer around tools, debris or equipment blocking their path. These obstacles must be detected in time for the control system to slow or stop the robot before a collision occurs. Close-range sensing helps support this movement between fastening tasks.

Multi-Point Ranging in Changing Light Conditions

Each NOVA20 sensor provides distance readings across 20 detection channels, covering multiple points within its field of view rather than relying on a single measurement point. The robot’s controller can use these readings to support decisions about when to slow down or stop.

Railway maintenance can take place on exposed outdoor tracks, in tunnels or at night. NOVA20 uses active laser ranging, enabling distance measurement in low light without relying on visible-light illumination. It also supports ranging up to 3 m under ambient light levels of 100 klx, enabling close-range obstacle detection in bright outdoor conditions.

Compact Sensors for Space-Constrained Robots

Drive components, fastening tools, batteries and control electronics compete for space within a rail-mounted robot. Obstacle-detection sensors must fit around these components while meeting coverage and power requirements.

Approximately the size of a beer bottle cap, NOVA20 weighs around 3 grams and consumes less than 1 W on average. Its compact form factor allows installation in confined locations, and its fully solid-state design has no moving parts.

NOVA20 outputs distance measurements for individual channels. The controller can use these readings for proximity checks without requiring image processing to obtain the distances.

The deployment brings fastening automation and close-range obstacle detection together on the same platform, supporting the use of robots in everyday railway maintenance.

About Benewake

Founded in 2015, Benewake is a provider of AI-native perception solutions for the physical AI era, with a mission to create the eyes that enable intelligent machines to perceive and understand the world. Guided by a model-defined hardware approach, the company continues to expand the capabilities of intelligent perception.

Drawing on full-stack expertise spanning optics, mechanical engineering, electronics, software, algorithms, and chip design, Benewake has built a comprehensive portfolio of “eyes for robots” across diverse applications. The company serves customers in more than 90 countries and regions, supported by a global network spanning R&D, manufacturing, sales, and engineering services.