Drone survey and mapping services, end to end

DGPS is differential GPS: a receiver on a known point broadcasts corrections to a roving receiver, cancelling most of the error common to both. In practice it takes satellite positioning from metre level to centimetre level. It is how a survey gets tied to a real coordinate system rather than floating on uncorrected GPS.

GNSS is the family of satellite constellations a receiver can listen to, including GPS, GLONASS, Galileo and BeiDou. DGPS describes the correction technique applied on top. So they are not alternatives: a modern survey receiver uses GNSS signals and applies differential correction, which is what brings the position down to centimetre level.

Yes, RTK is a carrier-phase form of differential correction, applied in real time from a base station or network. PPK does the same arithmetic after the flight instead of during it. Both resolve each photo centre or laser pulse to centimetre level before processing starts, rather than trusting the drone's onboard GPS.

The control report records where the control points were placed, how they were observed, which coordinate system and datum they are tied to, and what the independent checkpoints returned once processing was finished. That last part is the important one, because it states the accuracy actually achieved rather than the accuracy the method is capable of.

A contour survey records the shape of the ground as lines joining points of equal elevation. It is what drainage design, earthwork quantities and setting-out all start from, so an error here propagates into every downstream number. Contours are generated from a classified terrain model and delivered at whatever interval your design requires.

Grading and earthwork design mainly. The contour interval tells you how the water will run, where cut and fill will fall, and whether a proposed level works against the existing ground. On solar parks it drives the module table layout, and on highways it drives the cut and fill volumes that get certified.

That follows the terrain and the design. Flat ground usually needs a tighter interval to be useful, because a wide interval on a near-level site shows almost nothing. Steep ground can carry a wider one. Tell us what the contours are for and we will recommend an interval, then deliver at whatever you specify.

A topographic survey records the shape and features of a site: levels, breaklines, contours, and the existing structures and services visible on it. It is the base dataset a design is drawn on. Ours are delivered as contours in DXF, a terrain model, spot levels and a control report, written to your EPSG code.

A topographic survey describes the physical shape of the ground. A cadastral survey describes legal boundaries and ownership. They answer different questions and are often needed together, since a design has to sit inside the boundary as well as on the terrain. We can overlay cadastral information onto the topographic deliverable.

Photogrammetry turns overlapping aerial photographs into measurable geometry. It is the right tool for open ground, stockpiles, construction sites and anywhere you want a true-colour base map you can measure directly. It produces an orthomosaic, a dense point cloud and a textured 3D mesh from the same capture.

Aerial photogrammetry is photogrammetry flown from the air rather than shot from the ground. Images are captured with deliberate forward and side overlap so the same feature appears in several frames, and the geometry between those frames is what allows a measurable 3D model to be reconstructed from flat photographs.

Neither is better in general; they answer different questions. Photogrammetry gives true-colour, measurable imagery and works well on open ground. LiDAR pulses find gaps in vegetation and record returns from the ground beneath it, so it is the only aerial method that returns a true bare-earth surface under canopy. On mixed sites we fly both and merge them.

Whenever vegetation blocks the view of the ground. Forested ghat sections, reservoir slopes and overgrown corridors all return treetops from photogrammetry, which is the wrong surface for penstock alignment or reservoir capacity. We flew LiDAR for Central Railway across two ghat corridors and for EDF Energy at the Sawale hydro project for exactly this reason.

They reach a similar output by different means. 3D scanning, including LiDAR, measures distance directly with a laser. Photogrammetry infers geometry from overlapping photographs. Photogrammetry carries real colour information and is cheaper to fly; laser scanning penetrates vegetation and works regardless of light and texture.

A point cloud is the raw measured surface: millions of individual points, each with a coordinate and usually a colour or intensity value. It is what terrain models, cross sections and clash checks are derived from. We deliver it classified, as LAS or LAZ, so ground is already separated from vegetation and structures.

Classification tags each point with what it represents: ground, vegetation, building, and so on. It matters because the bare-earth surface used for terrain modelling is built only from the ground class. An unclassified cloud looks impressive and is difficult to design from, since treetops and rooftops sit in the same surface as the ground.

A digital terrain model represents bare earth with vegetation and structures removed. A digital surface model represents everything the sensor saw, including canopy and buildings. Earthwork and drainage design need the terrain model. Volumes over a stockpile or a progress comparison usually want the surface model. We deliver whichever the task needs, or both.

An orthomosaic is a set of aerial photographs corrected for terrain and camera geometry and stitched into one image tied to a coordinate system. Unlike a plain aerial photo it has consistent scale, so measurements taken on it are real. It is delivered as GeoTIFF and drops straight into GIS and CAD as a base layer.

Ground sample distance is how much real ground one image pixel covers. At 1.8 cm GSD, each pixel is about 1.8 cm across on the ground. It is set by flight height and sensor, and it sets the smallest detail that can be resolved, so it is chosen from the deliverable rather than from a default mission template.

Ground control ties the aerial capture to your real coordinate system, so the model sits where the site actually is rather than merely being internally consistent. Extra points are surveyed and then deliberately held out of processing as checkpoints. Comparing the finished model against those points is what produces an honest accuracy statement.

It maps the bed beneath a water surface. Water projects need the bed and the bank in one continuous model, because a reservoir capacity study, a drainage design or a dredging quantity all break down if the surface has a gap at the waterline. Aerial capture covers the land, bathymetry covers the submerged bed, and the two merge on a common datum.

Yes, and delivering them as one surface is the point. Aerial capture covers the land and the exposed bank, bathymetric data covers the submerged bed, and the two are merged on a common datum into a single terrain model. Volumes then compute across the full extent without a discontinuity at the shoreline.

Processing, classification, quality checking and delivery into the platforms your team already runs. Raw capture is not a deliverable; the value is in classifying the cloud, checking it against control, structuring the layers and handing over something that opens correctly first time. Output is written to your EPSG code, datum and layer convention.

Yes. Deliverables are written to the EPSG code, datum, units and layer convention you specify, so they drop into your existing project structure rather than needing rework on arrival. Tell us the coordinate system and the layer naming you use and it is set up before processing starts, not corrected afterwards.