Platform Overview

The complete drone data
processing pipeline

From processed capture to professional deliverable — the same QGIS/GDAL deliverables engine AeroLakeland has run on real commercial projects, now wrapped in a self-serve platform.

The Sample Site, five outputs: orthomosaic, surface model, 1 ft contours, point cloud and textured mesh. 32.7 acres mapped; the 9.2 million-point cloud is a LAZ download.

Architecture

Four-layer processing architecture

Each layer runs the same order every time — a real, repeatable pipeline, not a black box.

Ingest Layer

Accepts DJI Terra, Pix4D, Metashape, and WebODM exports (DJI Terra is the most-tested path). File-type and plan-limit checks on upload; coordinate-reference validation before processing.

Topographic contour plan of the Sample Site from its surface model, 1 ft interval with a 5 ft index; index contours labelled +4, +9, +14 and +19 ft, heights relative to the lowest 1 ft contour; a 200 ft scale bar and a north arrow. Marked up: spot elevations of +19.9 ft on the upper pad, +11.8 ft on the lower lot and +6.8 ft at the cul-de-sac, and a line from pad to lot, 253 ft long, a change of -3.8 ft, a 1.5 % grade

1 ft contours, 5 ft index · spots +19.9 / +11.8 / +6.8 ft · 253 ft line at 1.5 % grade

Deliverables Engine

QGIS/GDAL pipeline that turns your already-processed orthomosaic, DSM/DEM, point cloud, and 3D mesh into georeferenced KMZ, web-tiled overlays, and a branded PDF report (included with Pilot and Fleet, and during the free trial). Upload raw flight photos instead and a containerized OpenDroneMap stage reconstructs them first, then hands off to the same pipeline.

Elevation cross-check of the Sample Site surface model: a height distribution of every surface-model cell inside the mapped area, on an axis from 0 to 75 ft above the 1st percentile, with the 1st and 99th percentiles marked; 99.97 % of the mapped area has heights, 15 cm surface-model pixel, 5 cm orthomosaic pixel, a 73.0 ft spread between the 1st and 99th percentile, 5,873,939 cells checked. The surface includes trees and roofs; heights are relative, no vertical datum declared

Measured on the Sample Site: 5,873,939 cells checked · 99.97 % with heights · 15 cm surface / 5 cm ortho pixel

Automated QA

Tileset health checks and elevation cross-validation run on every job before results reach your portal.

The delivered Case Study #1 client portal open on the orthomosaic of the 32.6-acre materials yard, under its toolbar: compass views, distance, area, elevation, profile, volume, flood, sections, pins and a screenshot button

Case Study #1 · 32.6 acres at 1.7 cm per pixel · the portal as delivered

Deliverables Portal

Download files directly or share a revocable, expiring link with a client — no account required on their end.

Technical Specs

Built for professional accuracy

Results depend on capture method, overlap, and whether ground control points or RTK/PPK were used — these figures reflect real project output, not a lab benchmark.

Orthomosaic resolution1.7–1.8 cm/px on the case-study exports · 5 cm/px on the raw-photo case study at the fast preset — set by your flight and export
Elevation accuracyMeasured against the checkpoints you upload — reported as RMSE, mean and worst point — no pass/fail verdict
Typical turnaround24 hours — large or complex captures take longer
Supported exportsDJI Terra · Pix4D · Metashape · WebODM — included with Pilot and Fleet, and during the free trial full list
Output CRSAuto-detect WGS84, UTM, custom EPSG
Export formatsGeoTIFF · KMZ/KML · LAS/LAZ · OBJ · DXF · SHP · LandXML
Progress tracking2 to 8 flights per timeline — cut/fill/net per period, plus a weekly email how it works

Security & Access

Isolation at every layer

Data Encryption

TLS in transit. Files move via short-lived presigned URLs, never through a shared endpoint.

Isolated Tenancy

Per-account data scoping enforced at the database level on every request.

Role-Based Access

Owner, admin, and read-only roles enforced server-side, with every project scoped to its own account.

Controlled Sharing

Client links are revocable and can expire by date or view count.

Vertical accuracy

Every elevation, labelled with what it actually is

Ellipsoidal and orthometric heights differ by roughly 90 feet in Florida. Software that reports the datum you requested rather than the one your export actually carried will happily show you 90 feet of progress that never happened.

A → B, 580 ft across the Sample Site, 7.3 ft of relief along the line. Heights are relative to A because this sample declares no vertical datum, and the chart says so instead of printing a number that looks absolute.

Measured on every run

DroneFlow probes the vertical datum of every export it processes — not the setting you chose, the reference the data arrived on — and carries that verdict onto the project page, the comparison view, the cut/fill report and the AI assistant.

Mismatches are named, not averaged

If two flights sit on different references, the comparison says so rather than calling the difference progress. Upload surveyed checkpoints and it goes further: it tells you whether the error it measured is explained by the datum, or whether the problem is elsewhere.

An accuracy certificate

One page stating the delivered datum, the horizontal reference, the automated checks that ran, and — when you have supplied checkpoints — the measured error against them. It names what it does not verify, and it never grades the data. Whether the numbers meet your tolerance is your call.

Above-sea-level heights are not a North-America-only feature. Choose your vertical datum at upload — NAVD88 (GEOID18) for work specified against the US datum, or EGM2008 anywhere in the world. DroneFlow measures which datums actually resolve at your coordinates rather than assuming a region, and shows the offset each one applies at your site.

Works With Your Tools

Export formats your team already uses

No proprietary lock-in — every deliverable exports to a format that opens directly in the software you already run.

QGIS / ArcGIS

GeoTIFF, Shapefile-compatible exports

Google Earth

KMZ / KML with layer toggles

Point Cloud Viewers

LAS / LAZ dense point clouds

CAD & Civil

DXF contours, and LandXML surfaces for Civil 3D, Trimble Business Center and Carlson