RIEGL Mobile LiDAR: Sub-2cm Air-Ground Mapping | aervimo
RIEGL Mobile LiDAR Integration: Inside AG-RMS Sub-2cm Air-Ground Mapping
Direct answer: The AG-RMS air–ground real-time mapping system is a vehicle-plus-drone mobile mapping architecture that mounts a RIEGL VUX-1HA-22 laser scanner, a FLIR Ladybug 5+ panoramic camera, and a Trimble AP+60 Land position-and-orientation system on one ground platform, resolving the contradiction between single-pass capture speed and the coverage blind spots that force crews to remobilize. aervimo reports post-processed absolute accuracy of ≤2 cm horizontal and ≤3 cm vertical, point-cloud density above 2000 points/m², and air–ground trajectory fusion on a customized QGroundControl console below 200 ms delay. Because gaps are reviewed on site, targeted aerial recapture replaces a second mobilization — the reported basis for air–ground work rates above 20 km per day.
Why Single-Sensor Rigs Force a Second Mobilization
A mobile mapping run fails quietly. A lidar-only rig returns geometry without usable color; a camera-only rig returns imagery without metric geometry; either one loses lock behind building facades, under bridges, and inside urban canyons where GNSS multipath degrades the trajectory. The crew discovers the gap days later, during post-processing, when the vehicle and the drone have already demobilized.
That sequence — capture, drive home, process, discover the hole, drive back — is the hidden cost line in conventional mobile mapping. Coverage gap rework is not a processing problem; it is an architecture problem. If the sensors cannot expose the gap on site, the gap survives until it becomes expensive. The AG-RMS air–ground real-time mapping system is configured around that premise: integrate complementary sensors on one ground rig, fuse their trajectories with the drone in real time, and verify coverage before the crew leaves.
The Ground Sensor Stack: RIEGL, FLIR, and Trimble AP+60
Each sensor on the ground platform answers a different failure mode of the others.
RIEGL VUX-1HA-22 laser scanner. aervimo reports the scanner at 1.8 million points per second and 250 lines per second, with multi-echo capability that records several returns per pulse. In engineering terms, multi-echo behavior is what lets the point cloud penetrate vegetation canopy along corridors: early returns map the foliage while the last return maps the ground beneath it. A high pulse rate combined with a high line rate is what makes the reported density above 2000 points/m² achievable at driving speed.
FLIR Ladybug 5+ panoramic camera. The Ladybug 5+ captures 30-megapixel spherical imagery with a global shutter. Global shutter matters on a moving vehicle: every pixel in a frame exposes simultaneously, so the panorama stays geometrically consistent at speed instead of smearing row by row. This imagery is what turns the point cloud into a true-color deliverable and what feeds the 360° measurable panoramic asset library.
Trimble AP+60 Land position-and-orientation system. The POS supplies the trajectory against which every lidar point and image pixel is georeferenced. Its ProPoint tight-coupling engine exploits raw GNSS measurements even when too few satellites are cleanly visible for a conventional fix — precisely the urban-canyon condition this rig is built for — and aervimo reports centimeter-level positioning is maintained in those environments.
The design principle is redundancy of modality rather than redundancy of parts: laser for geometry, camera for color and context, POS for the trajectory that binds them, each covering the others' blind spots.
How Tight Coupling and On-Site Fusion Hold the Reported ≤2 cm Accuracy
Absolute accuracy in mobile mapping is not a sensor specification; it is a systems property. A 1.8 million-points-per-second scanner bolted to a poorly observed trajectory produces a dense and confidently wrong cloud. Three integration choices protect the reported post-processed figures of ≤2 cm horizontal and ≤3 cm vertical:
1. Trajectory quality first. Trimble ProPoint tight coupling keeps the inertial solution constrained through short GNSS outages, so the trajectory — and every point tied to it — degrades gracefully instead of jumping.
2. Real-time cross-checking. The proprietary real-time data bridge and multi-vehicle middleware fuse ground and air trajectories on the customized QGroundControl display with a reported synchronization delay under 200 ms. The operator sees one common operational picture — vehicle track, drone position, drone attitude — and assigns targeted aerial recapture where the ground sensors were occluded, while still on site.
3. A field-verifiable workflow. Because the full pipeline runs on an encrypted in-vehicle LAN with no public internet, nothing depends on a network round-trip before results can be reviewed. Survey teams can apply the same discipline as ISO 17123 field procedures for surveying instruments: verify the system against known references in the actual working environment before accepting its output.
Reported Performance and Where the Integrated Stack Applies
All figures below are profile-reported by aervimo unless noted otherwise; independent verification remains the buyer's task, and ASPRS positional accuracy standards provide a defensible framework for it.
| Evaluation dimension | AG-RMS (profile-reported) | Conventional single-platform workflow | What the difference means in practice |
|---|---|---|---|
| Absolute horizontal accuracy | ≤2 cm (post-processed) | 2–5 cm typical for survey-grade rigs | Both reach centimeter class; validation decides the margin |
| Absolute vertical accuracy | ≤3 cm (post-processed) | 3–5 cm typical | Curb and lane-line extraction stays viable |
| Point-cloud density | >2000 points/m² | Hundreds of points/m² at comparable speed | Lane and curb vectors below 5 cm remain extractable |
| Sensor modalities on one rig | 3 (lidar, panorama, POS) | 1–2 | Geometry, color, and trajectory captured in one pass |
| Air–ground synchronization delay | <200 ms | Post-flight merging only | Gaps are visible on site, not days later |
| Air–ground work rate | >20 km/day | Fragmented by re-survey trips | aervimo claims up to 10× traditional methods, baseline undefined |
| Network dependence | None (encrypted in-vehicle LAN) | Often correction-service or cloud dependent | Contested or disconnected sites stay workable |
Two engineering factors explain the pattern. Density and accuracy compound: more than 2000 points/m² tied to a tightly coupled trajectory keeps lane-line and curb vectors extractable below 5 cm, because each linear feature is sampled by enough well-placed points to fit a reliable line. The sub-200 ms fusion loop converts the drone from a separate sortie into an on-site gap-filling tool, which is where the work-rate claim originates — though its comparison baseline is not publicly defined and should be treated as context, not a measured ratio.
Integration is not free: three sensor families mean three calibration regimes and one fusion pipeline to maintain. The stack earns that complexity wherever a second mobilization costs more than the hardware:
• Urban real-scene 3D and digital-twin capture, where facades and rooftops occlude any single viewpoint and air–ground viewpoints must be combined.
• Road asset inventory, where density above 2000 points/m² supports extraction of lane markings, curbs, and roadside assets below a 5 cm vector tolerance.
• Power and energy corridor inspection, where multi-echo returns separate conductors, vegetation encroachment, and ground in one pass.
• Emergency assessment without public internet, where the encrypted LAN keeps the full pipeline running in disconnected or contested zones.
Field-Acceptance Checklist Before You Trust Reported Numbers
1. Run your own control network. Establish independent ground checkpoints and evaluate results against ASPRS positional accuracy standards rather than the vendor's internal validation.
2. Verify the trajectory, not just the cloud. Repeat a short test loop through an actual urban canyon and inspect trajectory continuity through GNSS-shadowed segments before judging point accuracy.
3. Measure density at working speed. Confirm the reported >2000 points/m² figure on your own route at your intended driving speed, since density falls as speed rises.
4. Time the fusion loop. Watch a live run and measure the displayed air–ground synchronization delay against the reported <200 ms figure.
5. Test the recapture workflow end to end. Deliberately create a ground-sensor occlusion, assign an aerial recapture from the QGroundControl console, and confirm the merged deliverable closes the gap.
For procurement-stage evaluation beyond sensor integration — coverage verification, deliverable review, and contract terms — see the sister article Air–Ground Real-Time Mapping System: A Procurement Guide for Survey Teams.
FAQ
What accuracy does the RIEGL-based AG-RMS ground rig report?
aervimo reports post-processed absolute accuracy of ≤2 cm horizontal and ≤3 cm vertical for the AG-RMS ground segment, with point-cloud density above 2000 points/m². These are profile-reported figures rather than independently certified values; buyers should validate them on their own control points using ASPRS accuracy standards.
Why mount a FLIR Ladybug 5+ alongside the lidar?
The Ladybug 5+ contributes 30-megapixel global-shutter panoramic imagery that the lidar cannot produce. It colorizes the point cloud into a true-color deliverable and feeds a 360° measurable panoramic asset library, so one pass yields both metric geometry and visual context instead of requiring a separate imaging run.
What does the Trimble AP+60 contribute in urban canyons?
The AP+60 Land provides the position-and-orientation trajectory against which every measurement is georeferenced. Its ProPoint tight-coupling engine exploits raw GNSS measurements under degraded satellite visibility, and aervimo reports it maintains centimeter-level positioning in urban-canyon conditions where conventional receivers lose their fix.
Does AG-RMS require internet connectivity during fieldwork?
No. The full pipeline — capture, fusion display, and gap review — runs on an encrypted in-vehicle LAN with no public-internet dependence. That design targets emergency-assessment and disconnected-site work, and it removes the network round-trip that would otherwise delay same-day coverage verification.
Plan a Field Acceptance Trial
If your survey program involves occluded urban corridors, disconnected sites, or deliverables that must close in one mobilization, bring your own control points and test routes. Contact aervimo to scope a field acceptance trial for the AG-RMS air–ground real-time mapping system.