20+ AI models.
One seamless pipeline.
This is not a simple drag-and-drop tool. Road Triage is a sophisticated ensemble of computer vision, machine learning, and pavement engineering models working in concert. It took years to build so that it runs automatically for you.
Reconstruction, not just detection.
Most pavement-assessment systems are detection systems with a reporting layer bolted on. Road Triage is a reconstruction system.
We combine imagery with motion signals to recover the camera's path through the scene, and we reconcile identified defects across multiple frames before they affect the score. The data we produce carries uncertainty information alongside it — so when the system isn't sure, it says so, instead of guessing.
17 distress types. Measured automatically.
Every image is analyzed by the Road Triage Perception Engine — our proprietary AI pipeline. Each distress is identified per ASTM E1778 terminology, classified by severity per ASTM D6433, measured from imagery per ASTM E3303, and crack widths quantified per AASHTO R 55. Scoring applied via ODOT DSE and Wisconsin PASER methodologies alongside D6433. Applied to both asphalt and concrete pavements.
How we measure crack severity and distress dimensions.
Multi-Frame Observation
Road Triage does not rely on a single image. The Perception Engine reconciles identified distresses across multiple frames as the survey vehicle passes — the same crack may appear in 5 to 15 consecutive images from varying distances and angles. This builds a measurement profile for an identified distress, cross-validates between frames, and identifies the optimal observation distance for more reliable results.
Pixel-Precise Measurement
Distress dimensions — length, width, and area — are derived from pixel-precise AI segmentation masks, not bounding box approximations. Each pixel maps to a known ground area through our AI-powered camera calibration system, which automatically calibrates after data collection using reference objects and spatial relationships in the survey imagery. No specialized setup by drivers — just mount the camera and drive.
Standards-Based Classification
The calibrated measurement pipeline delivers estimated distress area in square feet, length and width in feet, and crack width classification into the industry-standard Low, Medium, and High severity categories defined by ASTM D6433 and AASHTO R 55. Precise area measurement combined with standardized severity classification gives engineers both the quantitative data for density calculations and the categorical data required by scoring methodologies.
Built-In Confidence Scoring
Measurements throughout the pipeline carry an internal confidence score. When the system is uncertain about a classification, it flags the observation for quality review rather than guessing. The result is more reliable measurement data for network-level planning.
One pipeline. Five scores.
Identified distresses are classified per ASTM E1778 terminology, measured via the Road Triage Perception Engine, and crack widths analyzed per ASTM E3303 and AASHTO R 55. That unified pipeline then feeds all five scoring methodologies. The same standards inform each score — they just use the results differently. Applied to both asphalt and concrete pavements.
Pavement Condition Index
Pavement Surface Evaluation and Rating
Cracking Condition Index
Pavement Condition Rating
Surface Distress Index
All five assessments apply to both asphalt and concrete pavements. Distress detection, classification, and measurement are performed once through a unified pipeline, then each scoring methodology applies its own calculation procedure to the shared distress data.
The Road Triage Helm.
All results flow into a powerful web-based mapping and analytics platform. Interactive maps, change detection, AI-powered queries, and 13+ export formats. Tour the full platform →
Interactive Maps
11+ visualization types including PCI, PASER, distress density heatmaps, and asset inventory overlays.
Change Detection
Compare assessment cycles side-by-side. Automatically flags segments that improved, held steady, or declined.
AI-Powered Analysis
Ask questions in natural language. Get instant insights about your network conditions, trends, and data.
Federal & State Data Overlays
Crash records, traffic data, flood zones, bridge inventory, census demographics, and more — overlaid on your network.
Comprehensive Data Export
PDF, Excel, CSV, GeoJSON, Shapefile, KML, GeoPackage, DXF, HTML, JSON, XML, SQLite, JSONL. Ready for any GIS platform.
Actionable Condition Data
Detailed distress inventories, condition scores, imagery, and analytics — everything your team needs to make informed decisions.
See the technology in action.
Schedule a demonstration and walk through the platform with our team.