Keywords: railway automation, automated shunting, simulation-to-reality transfer, laser scanning, coupling detection, point cloud processing.

This study presents a comprehensive investigation into the transferability of a laser scanning-based coupling detection algorithm from simulation environments to real-world applications in automatic shunting. With the growing implementation of automation in rail transport systems, ensuring efficient and safe coupling mechanisms has become an important consideration. A data processing pipeline was developed within a Gazebo simulation framework to develop the detection system, followed by extensive testing and adaptation on data collected from a real-world shunting vehicle. The results demonstrate that while accurate recognition of coupler positions can be achieved, challenges persist in adapting algorithms due to discrepancies between simulated and actual conditions. The findings in this paper highlight the importance of model fidelity and parameter tuning in ensuring reliable performance under varying environmental factors. This research advances the understanding of automated coupling systems and contributes to the knowledge needed to bridge the gap between simulation and reality in the context of railway automation.

download the full-text of this paper (PDF, 3744 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description