AEC AI IndexReality capture means recording the existing condition of a site, building or asset in 3D, using laser scanners, drones, photogrammetry or 360° cameras. For AEC teams it feeds as-built documentation, survey, progress tracking and scan to BIM. AI is changing the slow part of the workflow: turning millions of points or thousands of photos into classified, usable models and CAD. This guide covers the reality capture tools for AEC listed in the AEC AI Index, including the hardware they pair with. Tool names link to their directory pages.
The reality capture workflow
- Capture: terrestrial laser scanners, mobile and SLAM scanners, drones, or 360° and phone cameras collect raw data.
- Register and process: scans are aligned and photos are turned into point clouds, meshes and orthomosaics.
- Classify and extract: points are labeled (ground, pipes, walls, poles) and features are extracted as linework or objects.
- Deliver: results go into CAD, BIM (Revit/IFC), GIS or a project platform for comparison against design and schedule.
AI mostly helps in steps 3 and 4, where manual work has traditionally taken the most time.
Capture hardware (not AI, but essential)
Most workflows start with established hardware. Terrestrial scanners include Leica Geosystems (RTC360, BLK), FARO Focus and Trimble X-Series Scanners. NavVis offers wearable and trolley mobile mapping, and Emesent Hovermap handles GPS-denied spaces such as tunnels and culverts. For aerial work, DJI Enterprise drones with lidar payloads, Wingtra fixed-wing VTOL and ROCK Robotic drone LiDAR are common choices. These are listed in Beyond AI.
Photogrammetry and drone mapping
PIX4D turns drone and terrestrial photos into orthomosaics, point clouds and 3D models. DroneDeploy combines drone and ground capture with AI analytics for construction and energy sites, and Propeller Aero focuses on earthworks volumes and progress. Looq AI takes an image-based approach with a handheld device, producing survey-grade 3D and CAD-ready linework without lidar or drone permits.
Scan to BIM and point cloud classification
Scan to BIM has traditionally meant modelers tracing point clouds by hand. Several AI tools now automate parts of it. aurivus and Pointly classify point clouds to speed up modeling, and Flai automates classification of aerial LiDAR survey data. Moving Atoms offers automated scan to BIM, converting LiDAR, SLAM or phone scans into editable Revit/IFC models. For civil and survey teams, Mach9 auto-drafts features from mobile and aerial lidar into CAD-ready deliverables. Expect to review and correct output: accuracy depends on scan quality, occlusion and how well the object types match what the model was trained on.
360° capture and construction progress tracking
360° cameras make it cheap to document a whole jobsite regularly. OpenSpace maps 360° captures to plans for progress tracking, Cupix builds 360° digital twins for construction and facility management, and Matterport captures existing buildings with AI that labels spaces and generates floor plans. Progress tracking tools compare capture against BIM and schedule: Doxel, Buildots (helmet-mounted 360° cameras), Reconstruct and Track3D.
Reality capture for infrastructure and utilities
For linear assets and networks, reality capture feeds digital twins and GIS. Neara and Sharper Shape build utility network twins from LiDAR and imagery, Nearmap provides aerial imagery and 3D capture with AI feature extraction, and Bentley Systems (iTwin / OpenSite+) supports infrastructure digital twins. For underground assets, see the guide to AI tools for utilities.
Choosing a reality capture setup
- Define the deliverable first: a progress report, a topo survey, a Revit model and a GIS layer need different accuracy and capture methods.
- Match accuracy to purpose: survey-grade work usually needs control points and scanner or lidar data; progress documentation often does not.
- Test AI classification on your own scans: results vary by site type, so pilot on a representative project.
- Check file formats: confirm exports (E57, LAS/LAZ, RCP, IFC, DWG) work with your CAD, BIM and GIS tools.
- Plan data storage: point clouds are large, so check cloud limits, retention and sharing permissions.
More resources
See every tool in Reality Capture and Digital Twins, hardware in Beyond AI, and the guide to AI tools for civil engineers.