CLOUDAR

HYBRID AI FOR SURVEYING

Two sensors.
One AI model.

Curbs, sidewalks, markings, poles and covers — extracted from LiDAR and orthophotos into coded CAD.

Both inputs reach one model to identify named features.

FILES IN. DRAWING OUT.

One site.
Every feature on its layer.

A LAS point cloud and a GeoTIFF orthophoto go in. A coded DXF comes out, in the survey’s coordinates.

Orthophoto of an arterial road and intersection in Brampton
InputOrthophoto + LiDAR · Brampton, Ontario
The same ground with Cloudar’s drawing: curbs, sidewalks, lane lines, crosswalks and stop bars
OutputCloudar drawing · DXF layers

FIVE FEATURES

Found, traced and coded.
In one pass.

All five, in detail

THE WORKFLOW

Fly. Extract.
Review. Export.

See the workflow
01 / FLY

Laser and camera, same ground.

An M350 flies the site with the Zenmuse L2 and P1. The point cloud and the orthophoto are registered to your survey control.

02 / EXTRACT

One model reads both.

Height and shape from the laser, paint and covers from the photo. Each feature is read from the sensor that shows it.

03 / REVIEW

Check it against your records.

Every feature carries its confidence. Cloudar reports what is confirmed, moved, missing and new.

04 / EXPORT

Files that open in your CAD.

A DXF on your layer standard, a feature table and a job report. It runs on a desktop in your office.

THE INPUTS

Two ways
to see the ground.

The orthophoto carries paint, edges and covers. The point cloud carries height, shape and structure.

THE TEAM

Built in Toronto.

Faraj and Ali Abidi founded Cloudar to take the manual tracing and coding out of survey drawings.

Meet the team
Faraj

Faraj

Co-founder
Ali Abidi

Ali Abidi

Co-founder