Pipeline and infrastructure construction logistics corridor monitoring
Sequential satellite acquisitions over a construction corridor reveal active work fronts, equipment concentrations and staging-yard activity without a single site visit. Useful for project-finance oversight and supply-chain scheduling where physical access is constrained.
Sensors
- Planet SuperDove: 3 m ground sample distance, 8 spectral bands, daily revisit at mid-latitudes. Sufficient to distinguish individual heavy vehicles and detect fresh soil disturbance along a right-of-way.
- Sentinel-2 MSI: 10 m resolution in visible and near-infrared bands, 5-day revisit at the equator (2–3 days with both satellites). Free archive from 2015. Best for corridor-scale change detection and vegetation clearance mapping over tens of kilometres.
- Maxar WorldView-3: 30 cm panchromatic, 1.24 m multispectral. Tasked on demand. Resolves individual pipe joints, crane booms and truck types. Latency from tasking to delivery typically 24–48 hours for routine collection.
- Airbus Pléiades Neo: 30 cm native resolution, stereo and tri-stereo capable. Useful for elevation modelling of spoil heaps and stockpiles alongside equipment counting. Revisit up to twice daily at favourable latitudes.
What a cleared right-of-way looks like from 500 km up
A pipeline or road corridor under construction leaves a distinctive spectral and geometric signature. Vegetation is stripped in a linear band, exposing bare mineral soil with high reflectance in Sentinel-2 bands 4 and 8A. That contrast against surrounding canopy is detectable at 10 m resolution within a single acquisition and measurable in near-real time as the clearance front advances.
At Planet's 3 m resolution, the same corridor reveals texture: spoil mounds, access tracks branching off the main strip, and the elongated shadows of pipe-handling cranes or excavators. None of this requires specialist optics. It requires consistent revisit and a systematic comparison between sequential acquisitions, which is where the analytic value sits.
Tracking construction tempo without setting foot on site
Construction tempo along a linear corridor is not uniform. Crews leapfrog between spreads, equipment clusters at active work fronts and disperses as sections complete. A single snapshot is almost meaningless. A time series is not.
Change detection between Planet SuperDove acquisitions taken 24 hours apart can identify which kilometre-posts are active, where equipment is massing ahead of a river crossing or a compressor station pad, and where the corridor has gone quiet. Normalised Difference Vegetation Index differencing between Sentinel-2 scenes quantifies cleared area per reporting period to within roughly one hectare at 10 m resolution. That figure is directly comparable to contractor progress claims.
Staging yards are particularly informative. Pipe sections laid out in parallel rows, heavy-lift cranes parked in sequence, and fuel bowsers clustered near a generator set all indicate an imminent active spread. At WorldView-3's 30 cm resolution, individual vehicles are identifiable by shadow length and plan shape. Manual counting is feasible for targeted areas; automated detection using convolutional neural networks trained on labelled vehicle imagery is the published standard for operational programmes and is accurate to within roughly 10–15 percent on mixed-equipment yards where shadows overlap.
Honest limits of the optical approach
Cloud cover is the primary constraint. Tropical corridors during monsoon season can lose 80 percent or more of usable optical acquisitions. Sentinel-2 and Planet both carry passive optical sensors; neither penetrates cloud. For corridors in persistently cloudy regions, optical time series must be supplemented by SAR, which is covered in the sibling page on container-yard monitoring and is not duplicated here.
Resolution sets a floor on what is detectable. At Sentinel-2's 10 m, individual vehicles are below the detection threshold. A cluster of ten or more heavy machines in a staging yard becomes a detectable radiometric anomaly but cannot be counted individually. Planet at 3 m resolves large equipment but cannot reliably distinguish a bulldozer from a water tanker. WorldView-3 at 30 cm resolves vehicle type but costs more per square kilometre to task and is not practical for continuous corridor-wide monitoring across hundreds of kilometres.
Revisit gaps matter for fast-moving projects. A spread advancing 1–2 km per day can move a full work front between acquisitions if cloud eliminates two or three consecutive passes. The practical approach is to use Sentinel-2 as the backbone for corridor-wide change detection, Planet for daily work-front localisation, and WorldView-3 or Pléiades Neo for targeted high-resolution verification of specific locations such as river crossings, compressor stations or border-crossing points.
Project finance and supply-chain applications
Lenders and equity investors in large infrastructure projects face a specific information problem: contractor progress reports arrive monthly, construction moves daily, and the assets are often in jurisdictions where independent verification is expensive or politically sensitive. Satellite monitoring provides an independent, time-stamped record of physical progress that is not derived from the contractor's own reporting.
The deliverable for a project-finance client is typically a corridor progress index: percentage of total right-of-way cleared, number of active spreads, location of each work front and comparison against the approved construction schedule. That index, updated weekly or fortnightly, gives lenders an early signal of schedule slippage before it appears in formal reporting.
For supply-chain scheduling, the same data answers a different question. If pipe-delivery trucks are not appearing at a staging yard on schedule, the satellite record shows it. If a spoil heap is growing faster than expected, a rock formation or unexpected soil condition may be slowing the bore. These are signals that procurement and logistics teams can act on before delays compound.
Archive depth and what it is good for
Sentinel-2's free archive runs from 2015. For a corridor that received planning approval years before construction began, that archive provides a pre-disturbance baseline: natural vegetation cover, seasonal reflectance cycles, pre-existing tracks and cleared areas. Distinguishing construction-related clearance from agricultural or logging activity requires that baseline. Without it, change detection produces false positives.
Landsat's archive extends to 1972 at 30 m resolution, which is sufficient for corridor-scale historical analysis. For a new pipeline following an older route, Landsat can document the original construction period and the revegetation trajectory after commissioning, providing a rehabilitation benchmark for environmental compliance monitoring.
Satellize runs analytics on open constellations including Sentinel-2 and Landsat and adds commercial tasking from Planet, Maxar and Airbus on client licence. The Tonga crop-estimation programme demonstrates the organisation's approach to applying change-detection methods on agricultural landscapes; the same spectral-change and object-detection pipeline transfers directly to linear infrastructure corridors.
Putting numbers on what you will actually receive
A typical corridor monitoring output for a 500 km pipeline project includes: a georeferenced change-detection layer updated on each cloud-free acquisition, a weekly progress report with kilometre-post-level activity status, a staging-yard equipment-count table for designated high-resolution collection windows, and an alert when a work front stalls for more than five consecutive acquisition days.
Accuracy figures from published studies on similar applications suggest right-of-way clearance mapping at Sentinel-2 resolution achieves better than 90 percent producer accuracy against ground-truth polygons in open and semi-open terrain. Dense canopy reduces that figure. Vehicle counting at Planet resolution using automated methods achieves roughly 70–85 percent accuracy depending on shadow conditions and equipment density. These are honest ranges drawn from the published literature, not claims specific to any single deployment.
Typical figures
| Spatial resolution (change detection) | 10 m (Sentinel-2), 3 m (Planet SuperDove) |
| Spatial resolution (equipment counting) | 30 cm (WorldView-3, Pléiades Neo) |
| Revisit frequency | Daily (Planet); 2–5 days (Sentinel-2, both satellites); on-demand within 24–48 h (WorldView-3, Pléiades Neo) |
| Spectral bands used | Visible (RGB), near-infrared (NIR), red-edge; NDVI differencing for clearance mapping |
| Minimum detectable cleared width | Approximately 20–30 m at Sentinel-2 10 m resolution; approximately 6–9 m at Planet 3 m resolution |
| Minimum detectable equipment cluster | 10+ large vehicles as radiometric anomaly at 10 m; individual vehicles at 3 m; vehicle type at 30 cm |
| Archive depth | Sentinel-2 from 2015; Landsat from 1972; Planet from approximately 2016 (commercial licence) |
| Cloud limitation | Optical only; persistent cloud (tropical monsoon) can eliminate 80 %+ of acquisitions in a season |
| Delivery formats | GeoTIFF change layers, GeoJSON work-front polygons, CSV equipment-count tables, PDF progress reports |
| Latency from acquisition to report | Typically 24–72 hours for automated change layers; 5–7 days for weekly narrative progress reports |
Analytics Satellize can run
| Right-of-way clearance progress index | NDVI differencing between sequential Sentinel-2 acquisitions; change classified against pre-disturbance baseline | Weekly GeoJSON polygon layer with cleared-area statistics per kilometre-post segment |
| Active work-front localisation | Binary change detection on Planet SuperDove time series using spectral angle mapping or band-ratio thresholding | Daily alert feed with GPS coordinates of active spreads and five-day stall detection flag |
| Staging-yard equipment count | Object detection on WorldView-3 or Pléiades Neo imagery using shadow-length and plan-shape classification; CNN-based vehicle detector trained on labelled heavy-equipment imagery | Tabular count report per designated staging yard, updated on each high-resolution tasking window |
| Schedule-vs-actual progress comparison | Cumulative cleared-area curve from satellite record compared against contractor-submitted construction schedule milestones | Fortnightly PDF report for project-finance lenders with variance flags and kilometre-post status map |
| Pre-disturbance baseline and environmental footprint | Historical Sentinel-2 and Landsat archive analysis to establish natural land-cover state before construction onset | GeoTIFF baseline land-cover map and total disturbed-area calculation for environmental compliance reporting |
| Spoil-heap and stockpile volume change | Stereo or tri-stereo elevation modelling from Pléiades Neo; differenced against baseline DEM to estimate volume accumulation | Monthly volumetric change layer and summary table for logistics and waste-management scheduling |
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.