Finding Surveyed Assets in AR: From Map Coordinate to Camera View

A map is excellent at showing relationships, but it still asks the operator to translate a north-up symbol into the scene in front of them. That translation gets awkward in a field of identical covers, along a long utility corridor, or when vegetation has changed since the asset was recorded. AR asset finding puts the project into the camera view: turn toward the pin, follow the distance, and walk back to the recorded location.

This is a different job from displaying a finished 3D model in AR. It does not require a processed cloud scene. The useful inputs already exist in a survey project: stored coordinates, the device's current position, gravity, and heading.

What the finder is solving

OpenSurvey3D converts the offset between the current position and each project asset into a local east–north–up frame. ARKit supplies a gravity-aligned camera frame and the device heading connects that local frame to the view. The result is a pin that moves with the camera instead of a bearing that the operator must interpret mentally.

Open one asset from its record to make it the target, or open a browsed project to see its nearby assets together. The target card reports relative direction, bearing, and distance while the other pins preserve context.

The 250-metre working horizon

OpenSurvey3D shows project assets within 250 metres of the current anchor. That is a deliberate field limit, not a statement that more distant coordinates do not exist. Small position and heading errors create increasingly large visual offsets as the projection gets farther away. A map is the better tool for reaching the general work area; AR becomes useful for the final approach.

If the operator walks far enough that the original local frame is no longer a good representation of the current surroundings, the view can be re-anchored at the new position. Re-anchoring refreshes the local relationship without changing the stored survey coordinates.

Find-mode is not stakeout

AR is intentionally the finder. Stakeout is the setting tool. If a valve, control point, boundary corner, or design feature must be accepted at a stated tolerance, stop following the camera pin and use the numeric stakeout and check-shot workflow.

The uncertainty stack stays visible

A sharp pin is not proof of a sharp position. Its apparent placement inherits several uncertainties:

• Current position. Phone GPS may be several metres off; an RTK fixed rover can bring the horizontal origin to centimetre class in good conditions.
• Stored asset position. The original observation cannot become more accurate simply because it is shown in AR.
• Heading. Nearby vehicles, fences, structures, and magnetic materials can disturb a phone compass.
• Offsets and height. Antenna, camera, pole, and feature-centre relationships still matter.
• Local AR alignment. Motion tracking can drift, especially in low light or visually repetitive scenes.

That is why OpenSurvey3D keeps current positioning and heading quality in the AR view. Better GNSS improves the position component; it does not remove every other source of error.

Where AR finding earns its place

• Returning to control points, inspection assets, valves, manholes, markers, and previous observation locations.
• Finding the correct feature among many similar objects without repeatedly rotating a map.
• Walking a project with a colleague or client while keeping recorded assets visible in context.
• Reaching the vicinity of a concealed or overgrown asset before switching to the appropriate locator or verification method.
• Opening the record, photos, and anchored 3D scan once the right asset is selected.

3D scanning and AR finding form one return workflow

A georeferenced scan records what the asset and its surroundings looked like. AR finding answers how to get back there. Because both belong to the same project asset, the operator can capture a LiDAR mesh, point cloud, or photo scan during one visit, return through the AR finder later, and reopen the attached evidence without reconstructing the relationship from filenames.

The positional quality is still governed by the capture anchor. For a deeper explanation, see Georeferenced 3D Capture on iPhone.

A practical return-to-asset sequence

1. Open the project and use the map to reach the work area.
2. Open the asset record and choose Find in AR, or launch AR from the project browser to see nearby assets together.
3. Confirm the current position source and heading-quality indicators before trusting the apparent direction.
4. Follow the target arrow and decreasing distance while watching the real environment, not only the screen.
5. Re-anchor after a significant walk or whenever the visual alignment no longer agrees with the surroundings.
6. At the target vicinity, inspect the record and its attached photos or 3D captures.
7. If the job requires setting or acceptance, switch to stakeout and record an independent observation.