Municipal asset managers rarely know exactly what they own. Streetlights, traffic signals, road signs, utility poles and bollards are added, replaced and decommissioned over decades with inconsistent record-keeping, leaving maintenance budgets allocated against outdated inventories and energy audits built on guesswork. Field surveys are slow, expensive and typically cover only a fraction of the network in any given year, meaning the gap between the register and reality widens continuously.
Satellite optical constellations flying at sub-50cm resolution can now resolve individual streetlight columns and pole-mounted assets along any road in a city. Multispectral and thermal-infrared bands distinguish luminaire types — sodium vapour, LED, fluorescent — by spectral signature, and night-time low-light imagery detects which units are actually illuminated during operational hours. Machine-learning pipelines run object detection across city-wide image mosaics and cross-reference outputs against existing GIS layers, flagging every asset that is missing from the register, misclassified or dark when it should be lit.
The operational result is a living asset register accurate to street-segment level, refreshed on a quarterly or seasonal cycle without deploying a single field technician for the initial baseline pass. Maintenance crews are dispatched only to verified faults. Energy efficiency programmes know which luminaire types remain in service and where. Capital replacement plans are costed against real counts rather than estimates. Cities that have run these programmes report inventory discrepancies of 15–30% against their prior registers — a gap that directly translated into wasted maintenance spend and missed carbon-reduction targets.
Frequently asked
Can satellite imagery actually resolve individual streetlight poles, or is it too coarse?
Modern commercial satellites such as Maxar's WorldView Legion and Planet's SkySat series deliver 0.3–0.5 m ground sample distance, which is sufficient to detect and classify standard utility poles and luminaire heads in open road corridors. Accuracy drops in narrow alleys or beneath dense canopy, where oblique captures or SAR-derived shadow analysis can help. AI pipelines applied to this imagery now regularly achieve 88–93% detection recall in urban test sites.
Why would a city bother with satellites when it can do a ground survey or use mobile LiDAR?
Ground surveys and mobile LiDAR produce highly accurate snapshots, but they are expensive (typically $0.50–$2 per asset for a full van survey), slow to repeat, and leave data stale within months as new assets are installed and old ones removed. A sovereign satellite programme offers economical repeat passes over the entire city — not just surveyed corridors — enabling continuous change detection rather than a once-per-decade inventory refresh. The two approaches are complementary: satellite change detection flags where ground crews need to go.
What does 'sovereign' mean here — does the nation need to own the camera in orbit?
Full sovereignty means the nation owns the satellite, the ground segment, and the data pipeline, eliminating the risk that a foreign vendor denies access, raises prices, or degrades resolution during a geopolitical dispute. A pragmatic intermediate step is a national ground station that receives and archives raw imagery from partner satellites under intergovernmental agreements, with domestically hosted processing. Either model is far preferable to a pure software-as-a-service arrangement where the government has no data rights and no continuity guarantee.
How does this application integrate with a city's existing GIS or asset management system?
Detected assets are exported as GeoJSON or CityGML 3.0 feature layers (OGC standard OGC 20-010), which ingest directly into mainstream GIS platforms such as ESRI ArcGIS, QGIS, or any ISO 19115-compliant spatial data infrastructure. Unique asset IDs link to maintenance management systems via standard REST APIs. The satellite-derived layer acts as a continuous audit layer that triggers alerts when the field register and the satellite-observed reality diverge.
How often does the satellite inventory need to be refreshed to stay operationally useful?
For a city with moderate asset churn (new subdivisions, road works), a monthly refresh is adequate to maintain an asset register within a 5% staleness threshold. High-growth cities or those running LED retrofit programmes may need weekly tasking during active construction phases. LEO nanosatellite constellations with daily revisit, such as those operated by Planet or BlackSky, make weekly-or-better cadence affordable at city scale.
Is nightside imagery useful — can satellites detect which streetlights are actually switched on?
Yes. NOAA's VIIRS Day/Night Band and experimental high-resolution nighttime imaging payloads can map radiance at luminaire level in large open areas, enabling cities to cross-check which assets in their register are actually emitting light and flag dark segments for maintenance. Resolution and cloud limitations remain, but nightside passes add a functional-status dimension that daytime optical imagery cannot provide.
What are the data volume and processing demands for a city of 1 million people?
A 300 km² urban footprint at 0.3 m GSD generates roughly 3–4 GB of raw imagery per pass. Monthly change-detection runs on a cloud or sovereign compute cluster typically complete in under two hours using containerised AI inference pipelines. Storage for a rolling 24-month archive of monthly passes amounts to approximately 100–120 GB — well within the capacity of a mid-tier national spatial data infrastructure.
How does this capability intersect with smart-city IoT sensor networks?
Satellite-derived asset inventories form the authoritative spatial backbone — the 'ground truth' register — onto which IoT sensor data (smart-meter consumption readings, fault alerts from connected luminaires) is anchored. Without an accurate spatial register, IoT dashboards lose coherence when assets are moved, replaced, or newly installed. The satellite layer provides the continuous reconciliation mechanism that keeps the IoT layer geographically honest.