Evaluation of Field Ballast Degradation at Ft. Campbell Army Installation Using Computer Vision and Comparison with Ground Penetrating Radar

Ballast aggregates play a critical role in railway track performance by supporting rails and crossties, distributing wheel loads, and enabling drainage. Monitoring ballast degradation is essential for maintaining track functionality. Traditional evaluation methods, such as visual inspection and sieve analysis, are subjective, labor-intensive, and limited in representing spatial variability. To address these challenges, a Ballast Scanning Vehicle (BSV) has been recently developed at the University of Illinois to automate ballast condition assessments using computer vision empowered with a deep learning image analysis algorithm. Field tests were conducted at the Ft. Campbell Army installation in Kentucky, where the BSV scanned open shoulders of railroad track with clean and degraded ballast layers at multiple depths using line-scan and area-scan cameras, and a 3D scanner. Fouling Index (FI) and Particle Size Distribution (PSD) were estimated from scanned images and validated against laboratory sieve analysis and Ground Penetrating Radar (GPR) results. The findings demonstrate the BSV's capability to accurately assess ballast degradation while highlighting discrepancies with GPR due to differences in methodology. This study underscores the BSV's potential as a practical and efficient solution for advancing the accuracy and efficiency of railway ballast evaluation.