Stadium and arena construction milestone tracking
Very-high-resolution optical imagery from Pléiades Neo, SkySat and WorldView-3 can verify discrete construction milestones against project schedules, supporting lender draw-down decisions with independent, dated evidence.
Sensors
- Airbus Pléiades Neo: 30 cm native resolution in panchromatic, 1.2 m multispectral; daily revisit over a fixed point at mid-latitudes using the four-satellite constellation. Stereo and tri-stereo modes yield digital surface models at roughly 50 cm vertical accuracy for earthworks volume estimation.
- Maxar WorldView-3: 31 cm panchromatic, 1.24 m multispectral, 3.7 m SWIR. SWIR bands help distinguish membrane materials and wet concrete from dry substrate. Stereo pairs support photogrammetric DSM generation. Revisit approximately 1 day at mid-latitudes.
- Planet SkySat: 50 cm panchromatic, 1 m multispectral. Up to 12 tasked collections per day over a single site are possible, making it the highest-cadence option for catching milestone slippage within a weekly reporting window. Less suited to photogrammetric volume work than Pléiades Neo or WorldView-3 stereo pairs.
- BlackSky: 50 cm panchromatic. Small-satellite constellation optimised for rapid revisit and same-day tasking. Useful for confirming activity status between higher-resolution collections, though stereo capability is limited.
The 30 cm threshold and what crosses it
Structural steel members on a stadium frame typically range from 200 mm box columns to lattice trusses spanning 60 to 100 metres. At 50 cm resolution, the roof truss grid is visible as a pattern but individual members blur together. Drop to 30 cm or better and the distinction between erected and missing bays becomes unambiguous. This is not a marketing threshold; it follows directly from the Nyquist sampling argument applied to objects whose minimum dimension is roughly 0.3 to 0.5 metres.
Pléiades Neo and WorldView-3 both sit at or below 30 cm in panchromatic mode. That puts discrete structural elements, including column lines, primary roof purlins and the leading edge of membrane installation, within the resolving power of a single tasked collect. Pitch preparation is easier still: the colour contrast between a laid synthetic surface and bare sub-base is detectable even at 1 m multispectral resolution.
Milestone verification that lenders can actually use
Project-finance draw-down schedules for large stadia typically tie tranches to named physical milestones: substructure complete, superstructure to eaves level, roof structure erected, roof membrane closed, playing surface installed. Each milestone has a contractual date and a surveyor's sign-off. The problem is that a site visit by a lender's technical adviser is expensive, slow and happens at a point in time that the contractor knows about in advance.
A satellite collect, tasked with 24 to 48 hours' notice, is harder to stage. The image is timestamped, georeferenced and archived. Comparing it against the milestone definition, for example the percentage of roof bays with membrane visible from above, produces a quantitative score rather than a narrative judgement. Crucially, the archive persists: if a dispute arises eighteen months later about when a milestone was actually achieved, the imagery is still there.
The honest limit here is cloud cover. A single collect on a cloudy day yields nothing. A monitoring programme that tasks every three to five days over a multi-month construction window will typically accumulate enough clear collects to bracket each milestone to within one reporting period, but this depends on the site's climatological cloud frequency. Sites in persistently overcast climates need a denser tasking cadence and acceptance that some gaps will remain.
Photogrammetric volume estimation for earthworks
Stadium construction almost always involves significant earthworks: bowl excavation, cut-and-fill balance across the site, and spoil management. Stereo or tri-stereo imagery from Pléiades Neo or WorldView-3 can generate a digital surface model at roughly 50 cm horizontal posting and, under good conditions, 30 to 50 cm vertical accuracy. Differencing two DSMs taken at different dates gives a volumetric change estimate.
The published method is standard photogrammetric point-cloud generation followed by surface differencing, the same principle used in quarry and opencast mining surveys. Accuracy degrades in shadow zones, which are common inside a partially roofed bowl, and in areas of uniform texture such as compacted gravel. A well-designed stereo collect schedules acquisition when sun angle is high enough to minimise shadow while still providing sufficient relief displacement for depth recovery. For a bowl roughly 200 m in diameter, a tri-stereo collect from Pléiades Neo can cover the full site in a single pass.
Volume estimates from satellite stereo carry an uncertainty of roughly 5 to 10 percent on well-defined stockpiles with clear edges. Cut-and-fill balance checks at that accuracy are sufficient to flag significant discrepancies against the earthworks programme, though they will not replace a licensed surveyor's sign-off for contractual purposes.
Revisit frequency and the slippage detection window
A monthly collect will tell you what the site looks like once a month. That is rarely enough. If a structural milestone slips by three weeks, a monthly image may show it as complete and the slippage is invisible. Catching milestone slippage within a reporting period, typically monthly or quarterly in project-finance contexts, requires collects at a cadence shorter than the slippage tolerance.
In practice, a weekly tasking rhythm using SkySat or BlackSky for activity confirmation, supplemented by fortnightly higher-resolution Pléiades Neo or WorldView-3 collects for detailed milestone assessment, gives a reasonable balance between cost and detection confidence. The weekly low-resolution collects flag whether significant activity is occurring; the fortnightly high-resolution collects make the milestone call. If the weekly collect shows no activity during a period when the programme says steel erection should be underway, that is an early warning worth acting on before the next draw-down request arrives.
What satellite imagery cannot settle
Imagery verifies what is visible from above. It cannot confirm material specification, weld quality, concrete strength or compliance with building regulations. A roof that looks closed from 500 km altitude may have unsealed penetrations, incorrect membrane grade or inadequate fixings. Satellite milestone verification is a complement to, not a replacement for, physical inspection by a qualified engineer.
Interior works are largely invisible. Fit-out of hospitality suites, installation of broadcast infrastructure, electrical and plumbing rough-in: none of these appear in nadir or oblique optical imagery of a roofed structure. The method is therefore most powerful during the structural phase, from groundworks through to roof closure, and less useful once the building envelope is complete.
Satellize's analytics workflow for projects of this type follows the same photogrammetric and change-detection methods described here, applied through a structured milestone-verification protocol. The team has applied comparable remote-sensing approaches to agricultural monitoring, including the Kingdom of Tonga crop-estimation programme, and the underlying image-differencing and classification pipelines transfer directly to construction contexts.
Setting up a monitoring programme before the first collect
The most common mistake is commissioning imagery after a dispute has already arisen. Archive imagery may exist, but tasked stereo pairs for volume estimation will not. Starting the monitoring programme at groundworks stage, with a baseline DSM collected before significant earthmoving begins, is the only way to get a defensible cut-and-fill record.
A well-structured programme defines milestone polygons in advance: the roof plane, the pitch rectangle, the perimeter road. Each collect is automatically clipped to those polygons and scored against a binary or percentage-complete criterion. Delivery as georeferenced GeoTIFF with an accompanying milestone scorecard in PDF or structured data feed gives the lender's technical adviser something to act on without needing GIS software. The key negotiation at programme setup is agreeing the milestone definitions with the borrower and contractor before construction starts, not after the first draw-down request.
Typical figures
| Best available panchromatic resolution | 30 cm (Pléiades Neo, WorldView-3) |
| Best available multispectral resolution | 1.2 m (Pléiades Neo); 1.24 m (WorldView-3) |
| Typical revisit at mid-latitudes | Daily or better for tasked VHR constellations; up to 12 collects per day with SkySat |
| Stereo DSM vertical accuracy | 30 to 50 cm under good conditions (Pléiades Neo tri-stereo, WorldView-3 stereo) |
| Volume estimation uncertainty | Approximately 5 to 10 percent on well-defined earthworks features |
| Minimum distinguishable structural element | Approximately 0.3 to 0.5 m linear features at 30 cm resolution |
| Spectral bands relevant to this use case | Panchromatic (structure); RGB and NIR (surface classification); SWIR (WorldView-3, membrane and moisture discrimination) |
| Tasking lead time | 24 to 48 hours for standard priority; same-day options available on some platforms |
| Archive depth | WorldView-3 from 2014; Pléiades from 2012; SkySat from approximately 2014 |
| Delivery formats | Georeferenced GeoTIFF, point cloud (LAS/LAZ), DSM raster, milestone scorecard (PDF or JSON feed) |
Analytics Satellize can run
| Milestone completion scorecard | Manual and semi-automated image interpretation against predefined milestone polygons; binary or percentage-complete classification | PDF report with annotated imagery and completion percentage per milestone, delivered within 48 hours of cloud-free collect |
| Earthworks volume change estimate | Photogrammetric DSM generation from stereo or tri-stereo pairs; surface differencing between dated epochs | Cut-and-fill volume table (CSV) with uncertainty bounds and DSM raster layers (GeoTIFF) |
| Roof membrane coverage percentage | Supervised pixel classification using panchromatic texture and multispectral or SWIR reflectance to distinguish membrane from bare steel or sky | Classified raster with coverage percentage and change map versus previous collect, in GeoTIFF and summary PDF |
| Construction activity status flag | Change detection between sequential collects; presence or absence of equipment, material movement and disturbed ground | Weekly activity status alert (email or API feed) indicating active, reduced or stalled construction |
| Pitch preparation stage classification | Multispectral classification distinguishing sub-base, root zone, and laid synthetic or natural turf by spectral signature | Classified polygon layer (GeoPackage) with stage label and collect date |
| Lender draw-down evidence pack | Compilation of milestone scorecards, annotated imagery and DSM-derived volume estimates for a specified draw-down date | Structured PDF evidence pack with image metadata, analyst certification and version-controlled archive reference |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.