Home / Projects

Complex data turned into safer, more resilient infrastructure.

We combine AI, engineering, remote sensing and advanced modelling, from disaster risk assessment to structural health monitoring.

100+Projects delivered
8Hazards modelled
25,000+Buildings assessed
Featured case studies
01 / 05
01
Florida, USHurricane

Hurricane-Induced Risk Assessment

An AI system analyzes aerial imagery to detect and quantify hurricane damage at the property level. It identifies damaged roof areas, rates their severity and maps impacts across whole regions.

Aerial imageryDamage detectionSeverity ratingRegional mapping

Who uses it Insurers estimating claims after landfall, emergency response teams prioritizing areas, and recovery planners.

Roof damage assessment dashboard on aerial imagery after a hurricane
02
California, USWildfire

Wildfire-Induced Risk Assessment

Machine learning assesses wildfire damage building by building from aerial imagery. Each building is classed as undamaged, partially damaged or fully destroyed, and the results feed interactive damage maps.

3-class damage modelAerial imageryInteractive maps

Who uses it Claims assessment, emergency response and rebuilding programmes.

Building-by-building wildfire damage classification map
03
Arctic regionPermafrost

Permafrost Damage Estimation

Predictive modelling combines satellite, terrain and environmental data to show where permafrost degradation threatens roads, airports and communities. The risk maps help prioritize monitoring, mitigation and long-term resilience.

Satellite dataTerrain modelsRisk mapping
Permafrost degradation risk map across Arctic roads, airports and communities
04
StructuralFire damage

House Fire Structural Damage Modelling and Mitigation

An assessment system for fire-damaged concrete members. It estimates fire exposure and remaining load capacity, grades damage severity and evaluates retrofit options such as CFRP strengthening.

Remaining capacitySeverity gradingCFRP retrofit
Fire-damaged concrete member assessment
05
StructuralDeep learning

Concrete Crack Identification, Classification and Measurement

A deep learning model built on a U-Net architecture detects, traces and classifies cracks in inspection images, then measures their length and width to support condition assessment and maintenance planning.

U-Net segmentationCrack length & widthInspection images
Concrete crack identification, classification and measurement workflow
Scroll to explore
Your project next

Have a structure or region to assess?