Showcase 01

Automated crack detection and classification

LCMSToolsPro automatically detecting and classifying nine distinct pavement defects on a single LCMS survey image. Each defect is labelled by type and severity: longitudinal, transverse, and alligator cracking, plus crack clusters. Bounding boxes are generated by our AI-powered analysis engine, with no manual review.

What used to take an engineer hours of image-by-image review now happens automatically. The system distinguishes crack orientation (longitudinal vs transverse), identifies fatigue cracking (alligator), flags crack clusters, and assigns severity levels (minor, medium, high) following the classification standard used in Brazilian pavement condition surveys.

Showcase 02

Surface depression detection with metric precision

Region-of-interest masking and metric measurement in the workbench. The engine detects surface depressions along the survey chainage, with a metre-scale ruler enabling precise location and sizing of each defect. Areas outside the lane boundary are masked to prevent false detections.

Beyond cracking, the workbench detects surface deformation such as depressions and rutting. The metric ruler along both axes lets engineers locate every defect precisely along the road, and the region-of-interest mask ensures the AI focuses only on the pavement surface, ignoring roadside artefacts and calibration bands.

The live workbench

A real, working product, not a concept

LCMSToolsPro runs today on real Brazilian survey data. The browser-based workbench combines AI defect overlays, a Laser QC audit panel, an annotation database, and GPS mapping in a single interface. Below is the Laser QC view auditing Delta Rutting points against AI-detected cracking on a live survey section.

Laser QC auditing 45 Delta Rutting points (15 to 50 mm) against AI-detected longitudinal, alligator and patch defects, with the live annotation database and tile analysis pipeline.

How It Works

Three Stages. One Integrated Pipeline.

STAGE 1

FIS to XML:
Raw Data Extraction

Processes raw .fis binary files from the official LCMS SDK (Pavemetrics LcmsAnalyser) and generates structured XML result files per road section. Supports batch processing of full survey campaigns with parallel execution and automatic IRI validation.

STAGE 2

XML to Structured Data Engineering

Parses all XML outputs into a normalised, queryable SQLite database. Captures cracking geometry, rut measurements, IRI values, GPS track, ravelling indicators, lane markings and pavement type classification. Built-in SQL query runner and Excel export.

STAGE 3

AI-Powered Visual Analysis

A browser-based engineering workbench where AI-detected defects are overlaid directly on LCMS images. Includes Laser QC panels for rut depth, delta rutting and roughness anomalies, GPS map integration, Google Street View cross-referencing, and audit trail for every annotation.

Technology

Built on Cloud Infrastructure and AI

LCMSToolsPro uses a multimodal AI engine as its core. Each LCMS image pair (Range + 3D) is fused into a composite RGB image and analysed by our AI engine with a specialised prompt encoding 13 years of Brazilian pavement inspection expertise. The system runs on Google Cloud infrastructure, with parallel processing of up to 16 concurrent analysis threads.

5-15x

faster than manual review

20x

defect detected

13

years LCMS expertise encoded

100%

ANTT and ARTESP regulatory alignment

WisePave works with road concessionaires across Brazil to turn LCMS survey data into regulatory-ready condition reports. Get in touch to see how LCMSToolsPro can reduce your survey analysis time and cost.

Bring AI-Powered Pavement Analysis to Your Network

© 2025. All rights reserved.