Data analytics occurs onboard in real-time within the RailAI™ Control System. Taking data from all onboard systems, it recognizes key elements of the track structure and compares current state against inspection limits from client-specific business rules. Exceptions to these rules are transmitted to the client as defects with type, severity, and precise location via cellular network within seconds of detection.
At a Glance
RailAI Control System
Automated collect-analyze-alarm on defect sequence
3D laser precision to minimize false alarms
Accuracy to readily locate defects for repair
The vast amount of data collected by the systems comprising RailAI are controlled and analyzed in real-time by the control system.
Essential measurements of track components that include crosstie condition, fastening systems, joint bars, welds, ballast profile, rail profile, and inventory are taken by Tetra Tech proprietary analytic software, then analyzed and compared against limits in the client’s business rules. Measurements outside these limits are reported in seconds with precise positioning data in a client-defined format for appropriate maintenance action.
Tetra Tech provides technical consulting services to customize RailAI artificial intelligence (AI) software to achieve customer-defined track assessment goals for both the real-time analysis and subsequent analysis of the full downloaded data set.
Monitoring and reporting in real-time is the desired goal but not at the cost of false alarms.
The 3D laser precision and ability to simultaneously process data from multiple systems provides a level of accuracy in identifying defects not available until RailAI. In addition to real-time processing, precise measurements are provided in elevation map and LiDAR point cloud formats, which can be reviewed after download, enabling post-processing and the ability to add new business rules on the fly as new track characteristics of interest are identified.