10.3.4 — EPC Monitoring — maturity: live

Stockpile Volume Estimation

Using satellite-derived stereo imagery and radar to measure the volume of bulk material stockpiles on EPC sites, providing independent quantity verification without boots on the ground.

Auto-drafted reference page — content reviewed for shape, individual references not yet link-checked.

On large EPC contracts — dams, refineries, highways, ports — the declared volume of aggregate, cement, steel, or earthwork spoil sitting on site is a direct proxy for schedule progress and contract payment entitlement. Contractors have every incentive to overstate stockpiles at payment milestones; owners and lenders have historically relied on expensive, infrequent surveyor visits that are trivially gamed. A sovereign infrastructure authority that depends on foreign EPC primes for national energy or transport assets faces the same information asymmetry on every project, compounded by the physical scale and access restrictions of active construction zones.

Satellite-derived photogrammetry closes that gap. Tasked very-high-resolution optical pairs (sub-50 cm GSD) generate dense point clouds over a site in a single pass; repeat-pass InSAR on the same area tracks cumulative earthwork movement between visits. On-board or near-real-time ground processing converts raw stereo imagery to a digital surface model, diffs it against the pre-construction baseline DEM, and outputs stockpile volumes per material class with typical accuracy of ±3–5% of true volume — well inside the ±10% tolerance accepted by most quantity surveyors. Revisit cadences of 3–5 days with a modest constellation make monthly surveyor visits look prehistoric.

For a government running a national infrastructure programme worth billions, the operational outcome is straightforward: every interim payment certificate can be cross-checked against an independently derived volume number before the treasury transfers funds. Disputes are resolved with photographic and geometric evidence rather than competing spreadsheets. And because the sensing and processing stack sits inside sovereign infrastructure, no contractor, no foreign imagery vendor, and no commercial due-diligence firm can see, delay, or manipulate the measurement before the government does.

What matters

Stereo optical point clouds achieve ±3–5% volumetric accuracy on bulk piles above roughly 500 m³ — within standard quantity-surveyor tolerance.
A 3–5 day revisit from a small constellation catches stockpile drawdown between milestones, exposing re-use of materials already paid for.
Baseline DEM acquisition before groundbreak is the single most critical data-collection act; missing it makes all subsequent differencing unreliable.
Export controls on very-high-resolution optical satellites (US ITAR, French MTCR commitments) mean access can be suspended by a foreign government at any time.

Sovereignty score: 7/10 — A government that relies on a commercial vendor to measure its own contractors' stockpiles has handed the referee's whistle to a third party with its own commercial relationships.

Commercial imagery vendors operate under export-control regimes (US ITAR, French dual-use law) that allow access to be restricted or delayed precisely when a geopolitical dispute with a contractor's home country is most likely.
Foreign due-diligence firms contracted to provide volume estimates have confidentiality obligations to their parent companies and repeat clients that may conflict with the sovereign owner's need for unfiltered data.
A sovereign pipeline processes baseline DEMs and change-detection results on national GPU infrastructure, preventing raw site geometry — which may include sensitive defence or energy facility layouts — from residing on foreign commercial clouds.
Lender conditions on sovereign infrastructure bonds increasingly require independent quantity assurance; owning the sensing stack fulfils that condition without creating a dependency on a single commercial provider that can reprice or withdraw the service.

Reference architecture

Payload: Very-high-resolution stereo optical imager, 40–50 cm GSD in panchromatic, 1.6 m multispectral, 12 km swath; secondary X-band SAR payload, 1 m stripmap resolution, 30 km swath, for cloudy-season and night-time passes
Bus class: ESPA-class microsat, 150–200 kg, 600 W payload power budget; pointing agility ±45° cross-track for stereo collection in a single pass
Orbit: Sun-synchronous LEO at 480–530 km, 6-satellite walker constellation, local solar time 10:30 descending node; 3–4 day median revisit at mid-latitudes, 1–2 day revisit in tropical construction corridors
Ground segment: 3-station national network (X-band downlink, S-band TT&C) co-located with major infrastructure ministry data centres; SatNOGS UHF/VHF backup for housekeeping telemetry; dedicated 200 Mbps ground-to-cloud pipe for L0 bulk transfer
Software & data
Latency
Cost band
Lead time

References

ISPRS — Volumetric Measurement of Stockpiles Using UAV and Satellite Photogrammetry — ISPRS commission papers document that dense stereo matching from sub-50 cm resolution imagery routinely yields surface models accurate to 3–8% of stockpile volume, depending on pile geometry and image overlap.
World Bank — Procurement Monitoring and Infrastructure Governance — The World Bank highlights independent quantity verification as a core anti-corruption safeguard on large civil infrastructure loans, noting that contractor-declared quantities are a primary vector for cost inflation.
ESA — Sentinel-1 InSAR for Surface Deformation and Earthwork Monitoring — ESA documents Sentinel-1 repeat-pass InSAR achieving millimetre-scale vertical displacement sensitivity, applicable to tracking earthwork volume change on active construction sites.
USGS — 3DEP Digital Elevation Program: Baseline DEM Standards — USGS 3DEP guidance establishes that a pre-event baseline DEM at 1 m or better resolution is the prerequisite for any reliable volumetric change-detection analysis.
Planet Labs — Monitoring at Scale: Daily Satellite Imagery for Infrastructure — Planet's published construction use cases demonstrate that frequent revisit optical imagery enables automated progress tracking on large sites, though sovereign access to raw data and processing pipelines is not guaranteed under commercial terms.