Automated, Drone-Based Grade Crossing Inspection

Drone–based assessment of rail grade crossings offers a great potential for rapid assessment of hang-up condition, visual sight line analysis, and crossing infrastructure. Through the Phase I SBIR, a proof-of-concept system was demonstrated at highway-rail crossings near Chelsea, MI with partner review. The collected imagery was processed through 3D photogrammetric software to produce high resolution digital elevation models (DEMs). It was demonstrated how the DEMs could be used for analysis of humped crossings and visual sight lines through a customized ArcGIS Pro workflow. Automated sign identification was demonstrated through use of an online machine learning tool. The Phase II SBIR project seeks to fully automate the Railroad Grade Crossing Viewshed and Automated Profile Assessment Tools as documented in the Phase I final report. With an emphasis on rapid drone data collection and the use of high accuracy GPS, MTRI Inc. seeks to provide commercially-available, timely, and accurate inspection of rail grade crossing and surrounding areas. The addition of a secure website will enable a user-friendly method for customers to request crossings to be inspected, send data to be analyzed by MTRI Inc., and receive all final products and reports based on the results of the automated tools.