TL;DR:
- The LiDAR sensors on iPhone 12 Pro+ and iPad Pro M-series devices generate real-time depth maps that significantly improve AR plane detection, occlusion, and object placement compared to camera-only AR
- Apple’s RoomPlan API automates floor plan generation — walk around a room holding your phone and get a structured architectural model of walls, windows, doors, and furniture in under two minutes
- For professional 3D capture, apps like Polycam and 3D Scanner App export point clouds, meshes, and floor plans in formats (OBJ, USDZ, E57, DXF) ready for use in CAD, BIM, game engines, or XR experiences
The LiDAR scanner that arrived quietly on iPhone 12 Pro in 2020 is now six hardware generations old, and it’s still underused — mostly because the consumer use cases Apple marketed (faster AR animals in a park) undersold what the sensor actually enables. For people building spatial computing applications, processing physical environments for digital twins, or capturing spaces for architectural and real estate use, an iPhone or iPad Pro is a surprisingly capable scanning platform.
This guide covers what the LiDAR sensor does, how to get the most out of it for different use cases, and which apps to reach for.
How LiDAR Works on iPhone and iPad
Traditional phone cameras infer depth from visual cues — texture, parallax, focus distance. The LiDAR scanner works differently: it fires a grid of invisible infrared laser pulses and measures the time it takes for each to reflect back. This time-of-flight measurement produces a dense depth map of the scene independent of lighting conditions or surface texture.
The effective range on iPhone LiDAR is approximately 5 metres for high accuracy, extending to around 10 metres for rough depth estimation. This is appropriate for interior room scanning but not large outdoor environments or distant objects.
The depth data is fused with camera imagery and IMU data in Apple’s ARKit framework, which combines the LiDAR point cloud with visual features for a more complete scene understanding.
Use Case 1: Better AR Experiences
The most immediate benefit is in ARKit-based applications. With LiDAR, ARKit can:
Detect surfaces in milliseconds: Camera-only AR typically takes several seconds to detect a flat surface. LiDAR detects floors, walls, and tables almost instantly, making object placement responsive.
Real occlusion: Without LiDAR, a placed 3D object appears in front of everything in the scene, breaking the illusion when a person walks between the camera and the virtual object. LiDAR depth data enables occlusion — real objects in the foreground correctly overlap virtual objects behind them.
Mesh generation: ARKit’s Scene Reconstruction feature (requiring LiDAR) generates a real-time polygon mesh of the entire visible scene, enabling physics interactions — a virtual ball that rolls off a real table and falls to the real floor, for example.
If you’re developing ARKit apps and targeting iPhone 12 Pro or newer, enabling these features is as simple as setting ARWorldTrackingConfiguration.sceneReconstruction = .meshWithClassification. The classification flag adds semantic labels to mesh surfaces (floor, wall, ceiling, table, seat, etc.).
Use Case 2: Floor Plans with RoomPlan
Apple’s RoomPlan API, available since iOS 16, uses LiDAR plus machine learning to generate structured floor plans automatically. You walk around a room holding your phone, and RoomPlan identifies:
- Walls, windows, and doors with dimensions
- Furniture (sofas, tables, beds, shelves) as parametric 3D objects
- Room connectivity (which doorways connect which spaces)
The output is a CapturedRoom structure that you can export as a USDZ file or process programmatically. Apple added multi-room merging in later iOS versions, enabling whole-apartment floor plans.
For real estate, property management, and facilities management use cases, this is a genuinely useful capability. An estate agent with an iPhone can generate a dimensioned floor plan in two minutes that would previously have required a measuring wheel, a tape measure, and manual drawing.
Accuracy is good enough for most commercial purposes — typically within 1–5% of actual dimensions for rooms up to 6x6 metres. For precision architectural or structural work, dedicated laser scanners (Leica BLK, Matterport Pro3) are more accurate, but they cost thousands of pounds.
Use Case 3: Point Cloud Capture with Polycam
For higher-quality 3D capture — objects, rooms, or building facades — Polycam is the most capable iOS scanning app and the one most widely used by professionals.
Polycam has two scanning modes:
LiDAR mode: Fast, real-time scanning that generates a point cloud from LiDAR data as you move around the subject. Good for rooms, architectural interiors, and scenes where you need quick turnaround. Output: OBJ, USDZ, GLTF, XYZ point cloud, or floor plan PDF.
Photogrammetry mode: Slower, takes dozens of photos and processes them via photogrammetry (either on-device or in the cloud) to generate a textured mesh. Better for detailed objects where surface texture and colour accuracy matter. Requires good lighting.
For room scanning:
- Open LiDAR mode
- Move slowly around the room, pointing at all surfaces
- Allow 3–5 minutes for a standard bedroom or living room
- Export as OBJ for import into Blender, SketchUp, or Unreal Engine
For architecture and construction: the E57 point cloud format exports from Polycam and is compatible with AutoCAD, Revit, and most BIM platforms.
Use Case 4: Digital Twins and Asset Documentation
In industrial and commercial settings, LiDAR scanning with iPad Pro is increasingly used for:
As-built documentation: Capturing the current state of a facility, plant room, or mechanical space for comparison with design drawings. A scan before and after renovation work creates a documented record.
Equipment inventory: Scanning a server room, warehouse racking configuration, or equipment layout creates a spatial record that’s easier to navigate than photos alone.
XR training content creation: Scanning a physical workspace to create a digital replica for AR-based training applications. Workers train in a virtual copy of the actual environment.
For these workflows, Reality Composer Pro (free from Apple) can import USDZ scans and add interactive elements, making scanned environments directly usable in visionOS or ARKit applications.
Limitations to Know
Reflective and transparent surfaces: Glass, mirrors, and highly polished metal don’t reflect infrared reliably. LiDAR struggles with windows (produces holes in the mesh), glossy flooring (inaccurate depth), and glass cabinets.
Outdoor range: At 5–10 metres effective range, LiDAR is not useful for façade scanning of multi-storey buildings or large outdoor structures. Photogrammetry via drone is better for those scenarios.
Occlusion in capture: The LiDAR scanner only captures what it can see. Tall furniture against walls, behind-sofa zones, and recessed areas require deliberate coverage — moving to get line of sight rather than just circling the room.
File size: Dense LiDAR point clouds of large spaces get big. A 60-square-metre flat scan in Polycam can produce 200–500MB files before processing.
The combination of LiDAR accuracy, ARKit depth integration, and apps like Polycam and RoomPlan has made consumer-grade spatial capture genuinely useful in ways that weren’t viable two years ago. If your XR projects involve capturing physical environments, the iPhone in your pocket is often the fastest way to start.