# aervimo engineering — technical reference Source: Technical Capabilities — External Presentation (技术能力介绍-对外展示版(6).pdf), 2026-08. Reviewed 2026-09-15. Figures retain their source-specific scope. This reference and the visible website are generated from the same content. ## AG-RMS air–ground real-time mapping system /air-ground-mapping/ AG-RMS combines vehicle and drone capture on a customized QGroundControl display. The profile reports <200 ms synchronization, absolute positioning accuracy ≤2 cm, elevation ≤3 cm and operation over an encrypted in-vehicle LAN. AG-RMS brings ground-vehicle and aerial viewpoints into one mapping workflow. A proprietary real-time data bridge and multi-vehicle communications middleware synchronize positions, attitudes and trajectories on the vehicle’s QGroundControl display. Operators review occluded areas and coverage gaps while still on site, then assign targeted aerial recapture. - Synchronization delay: <200 ms. Profile-reported synchronization of air and ground information on the shared display. - Absolute positioning accuracy: ≤2 cm. Source label: absolute positioning accuracy. Elevation is listed separately at ≤3 cm. - Points per m²: >2000. Reported point-cloud density for the described AG-RMS configuration. - Air–ground work rate: >20 km/day. Reported linear survey work rate; actual route and capture conditions are not specified. - Elevation accuracy: ≤3 cm. Separate source-listed elevation figure; validate against independent survey control. - Single-pass coverage claim: 100%. Source-stated workflow claim. Required surfaces and trial denominator are not defined, so this is not a universal coverage guarantee. - Comparison with traditional surveying: 10×. Reported comparison figure; the baseline and calculation method are not provided. - Shorter project duration: 80%. Separately reported duration reduction. Do not treat this and the 10× figure as mathematically equivalent. ### A common operational picture Display vehicle trajectories and drone positions and attitudes together, giving the operator a view of captured areas and places needing another viewpoint. ### Target gaps before demobilization Use aerial capture to complement ground-level blind areas. The intended value is to discover and fill omissions during the same mobilization. ### Work over the local vehicle network The described configuration transfers data through an encrypted in-vehicle LAN and does not require an internet connection for that workflow. ### Reference hardware and software configuration | Component | Model / platform | Source-listed role or rating | | --- | --- | --- | | LiDAR | RIEGL VUX-1HA-22 | 1.8 million points/s | | Position and orientation | Trimble AP+60 Land | Positioning and attitude reference | | Panoramic imaging | FLIR Ladybug 5+ | 30 megapixels | | Aerial platform | DJI M350 RTK | Aerial capture platform | | Common display | Customized QGroundControl | Air–ground positions, attitudes and trajectories | | Data and communications | Proprietary real-time bridge + multi-vehicle middleware | Synchronized field workflow over an encrypted in-vehicle LAN | This is the configuration identified in the supplied profile, not a claim of universal sensor compatibility. ### Smart cities and real-scene 3D — Application context Combine ground and aerial viewpoints for urban 3D reconstruction, reviewing occluded surfaces before the crew leaves. ### Road asset inventory and maintenance — Application context Capture road corridors and surrounding assets with a common coverage view, then target areas missed by the vehicle. ### Power and energy corridors — Application context Coordinate complementary observations of corridor assets and surrounding ground features for inspection and mapping. ### Emergency assessment — Application context Support civilian mapping and coverage review when public internet connectivity is unavailable. ### Development status and scope The profile presents AG-RMS as a core integrated product and illustrates a vehicle platform and 3D reconstruction results. Project deployment starts with sensor compatibility, route scope and field acceptance criteria. The figures above are reported in the supplied profile. Measurement conditions and full benchmark records are not supplied. Synchronization, positioning accuracy, cloud quality and coverage need separate acceptance criteria. Source pages: 12. ## ZhiTu VisNav map-referenced visual navigation /visual-positioning/ Map-referenced visual navigation using a down-facing camera and edge compute. A reported 740-frame campus trial achieved a median position deviation of 9.8 m at 120 m altitude, with an unoptimized prototype time of 0.23 s/frame. ZhiTu VisNav compares onboard camera imagery with a preloaded satellite or aerial reference map to produce an absolute visual position observation. The concept combines inertial information, learned reference-image matching and a quality gate. The profile describes a real campus trial and ongoing migration to an embedded platform. - Median position deviation: 9.8 m. Reported at 120 m altitude in the campus trial; this is a median, not a maximum-error bound. - Consecutive localization frames: 740. Profile-reported closed-loop sequence; not 740 independent flights. - Processing time per frame: 0.23 s. Unoptimized prototype measurement; not a released embedded-platform latency specification. - Reference-map area: ~1300 × 1300 m. Campus environment used for the reported evaluation. - Reported reacquisition after map exit: 100%. Observed in the described sequence, with no loss or divergence reported; the number of reacquisition events is not given. - Largest cited false-position error: 430 m. The quality gate reportedly rejected high-confidence false matches up to this deviation. - Reported false-match rejection: 100%. Reported for the described test; the number of false-match events and broader operating-domain statistics are not provided. ### Camera and reference map A down-facing camera and edge-computing board use a preloaded satellite or aerial map as a reference for absolute visual positioning. ### Complementary information sources The public architecture combines inertial information, reference-image matching and quality checks to assess whether a visual position observation is usable. ### Quality as part of the result The profile reports rejecting confident but incorrect position matches. Evaluation therefore includes false matches and availability alongside typical position error. ### Reported campus evaluation | Item | Condition / observation | | --- | --- | | Environment | Real campus scene with approximately 1300 × 1300 m reference map | | Altitude | 120 m for the cited median-error result | | Sequence | 740 consecutive localization frames | | Position deviation | 9.8 m median | | Prototype timing | 0.23 s/frame, unoptimized | | GPS use | First-frame initialization; the profile states GPS was not used for subsequent position estimates | | Reference comparison | The profile shows GPS ground truth compared with visual estimates | | Reacquisition and false matches | 100% figures are sequence-specific; event counts are not disclosed | ### Engineering maturity | Stage | Status described in the profile | | --- | --- | | Campus validation | Completed prototype sequence reported | | Rockchip RK3588 migration | In progress | | NPU quantization and C++ pipeline engineering | In progress | | Real-time platform interface integration | In progress | ### Civilian mapping-platform research — Application context Evaluate a map-referenced position observation in controlled conditions when conventional positioning is unavailable or unreliable. ### OEM feasibility assessment — Application context Review map dependence, camera conditions, latency, typical error and incorrect matches before defining an integration program. ### Reference-map suitability studies — Application context Compare the intended operating environment with the reported campus scene and identify the additional evidence needed. ### Development status and scope This is a campus-tested prototype undergoing engineering migration to Rockchip RK3588. The profile lists NPU quantization, C++ pipeline refactoring and real-time interface integration as ongoing work, not a released production SDK. All reported accuracy, timing and recovery figures refer to the described prototype test. They do not establish all-weather operation, unlimited drift-free performance, precision landing or deployment readiness. Source pages: 13.