From Point Clouds to Predictive Maintenance: A Review of Intelligent Railway Infrastructure Monitoring.

Area of Science:

• Geospatial Engineering
• Civil Engineering
• Computer Science

Background:

• Point cloud technology offers high-precision 3D geometric data crucial for complex railway environments.
• Intelligent detection, monitoring, and maintenance of railway infrastructure rely heavily on this technology.

Purpose of the Study:

• To systematically review point cloud applications in critical railway scenarios.
• To evaluate acquisition platforms, processing workflows, and advanced algorithms.
• To explore the integration of point clouds with computational mechanics for digital twins and simulations.

Main Methods:

• Systematic literature review of point cloud applications in railway infrastructure.
• Evaluation of mobile and stationary acquisition platforms and data processing workflows.
• Synthesis of advancements in feature extraction, semantic segmentation, deep learning, AI, and data fusion.

Main Results:

• Point clouds are applied in track geometry, component identification, tunnel/bridge modeling, clearance analysis, and condition monitoring.
• Deep learning and AI significantly impact data fusion and analysis.
• Synergy with computational mechanics enables high-fidelity digital twins for real-time simulations.

Conclusions:

• Persistent challenges include acquisition efficiency, precision, data fragmentation, and multi-modal fusion.
• Future research should focus on autonomous sensing, multi-sensor integration, and digital transformation for sustainable intelligent railways.