With the advancements in rail transportation in recent years, integrating fully automated operation modes into railway systems is becoming more prevalent. Thus, there is an urgent demand for detecting potential obstacles along railway tracks, such as pedestrians, animals, and falling rocks.
LiDAR technology has emerged as a crucial sensor with its exceptional obstacle-detection capabilities, precise detection accuracy, and adaptability to different environmental conditions. LiDAR collects echo signals from obstacles with laser pulses, and output 3D point cloud that outline detailed information about surrounding objects.
Benewake has been deeply involved in the field of railway transportation system. Based on its 3D LiDAR technology, Benwake offered a Meli train collision avoidance system, which could be integrated into signaling system, and help the achievement of automatic train operation.
Relocation based on HD map in GNSS denied environment With a multi-sensor fusion technology, Meli system could achieve relocation of ego train in the GNSS denied environment like tunnel, which also a challenge due to its less feature environment. The Accuracy of relocation could be less than 20cm, as calculated in platforms. Which is significantly improvement based on GNSS technology. | Free space planning and recognition based on HD map Free space planning could be done based on HD map, with the features of railway track. Scenarios in railway transportation like rolling stock depot, platform, intersection, tunnel could be accurately recognized, thus the relocation error could be remend by compensation algorithm. | |
Obstacle detection &recognition Multi-sensor fusion technology is applied in Meli system for obstacle detection, which improves the system perception ability. As in experiments, meli perception system could achieve the obstacle detection at •Train within 280m •Pedestrian within 200m •Small objects at size of 30*30cm within 80m | Traffic light recognition and traffic light color classification Traffic light recognition and color classification are critical to make sure the driving safety and efficiency in complex rail tracks. Technology of computer vision, machine learning, and sensor fusion is applied in this function. |