Arms-control and treaty verification

States that rely on partners or the IAEA to verify compliance are, functionally, trusting declarations they cannot check. A national technical means programme changes that equation with owned sensors, owned data and defensible evidence.

The dependence this ends: Trusting declarations you cannot check

The verification gap that most governments quietly accept

Treaty compliance has always depended on a combination of declared information, inspection regimes and intelligence. For most states, the intelligence leg of that tripod belongs to someone else. A government that cannot independently image a declared storage site, detect unreported construction or measure atmospheric signatures from a suspected facility is, in practice, delegating its national security judgement to a partner whose interests may not be perfectly aligned.

The commercial satellite record has already demonstrated what is possible without classified systems. Analysts using Planet, Maxar and Capella imagery have publicly documented missile-base construction in China, centrifuge hall activity in Iran and grain-silo expansion in contested regions, all from open-source data. That precedent matters: it establishes the evidentiary floor that a dedicated national programme can exceed, with controlled access, chain-of-custody metadata and no shutter-control risk from a foreign operator. Post-2022, the risk of commercial providers restricting imagery over politically sensitive areas is no longer theoretical. Owning the sensor is the only way to own the evidence.

What verification actually requires from a sensor mix

No single sensor type is sufficient. High-resolution panchromatic optical imagery resolves vehicles, construction activity and facility configuration changes at sub-metre scale, but cloud cover and night-time operations defeat it. X-band synthetic aperture radar operates through cloud and darkness, detects surface changes at centimetre-level displacement using coherent change detection, and can identify disturbed earth consistent with excavation or burial. Together, optical and SAR provide the temporal and environmental coverage that single-modality systems cannot.

The third sensor class is less discussed but increasingly important: atmospheric sounders and gas spectrometers. Instruments in the SWIR and TIR bands can detect methane, nitrogen dioxide, sulphur hexafluoride and other gases associated with specific industrial or weapons-related processes. SF6, for example, is used in high-voltage switching gear and certain nuclear applications; anomalous concentrations near a declared civilian site are a legitimate analytical flag. These detections are probabilistic, not conclusive, and atmospheric dispersion modelling introduces uncertainty that must be stated honestly in any evidentiary product.

Evidentiary standards matter as much as sensor capability. Data that will be presented to a treaty body, a UN panel or a national legislature needs documented provenance: satellite ephemeris, sensor calibration state, processing chain and analyst methodology. In-country processing facilities with air-gapped networks and access logs satisfy that requirement in a way that cloud-based third-party pipelines cannot.

The ambition ladder: pathfinder to operational constellation

A pathfinder programme typically involves one or two satellites, a single ground station and a small analyst cadre. The objective is to establish the legal, technical and operational baseline: ITU filing, frequency coordination, sensor calibration against known reference sites, and the development of an evidence-production workflow that meets the state's own judicial or diplomatic standards. Small-satellite missions of this class, with a 50-to-150 kg bus and a sub-metre optical or X-band SAR payload, have publicly reported budgets in the low tens of millions of dollars. Timeline from contract to on-orbit is typically 24 to 36 months for a first satellite, depending on payload heritage.

An operational constellation adds revisit. A single satellite in a sun-synchronous orbit at 500 to 600 km altitude passes over any given mid-latitude point roughly once per day. Two satellites on complementary planes halve that. Four satellites, with a mix of optical and SAR payloads, can achieve same-day revisit over priority sites and provide the cross-modal corroboration that strengthens an evidentiary case. Constellation programmes of this scale are analogous in scope to national Earth-observation programmes such as South Korea's KOMPSAT series or Israel's OPTSAT-3000, both of which are publicly documented. Budgets for multi-satellite national programmes of that heritage class run into the hundreds of millions over a decade, inclusive of ground infrastructure.

The honest constraint is revisit versus coverage. A four-satellite constellation watching fifty declared sites globally will collect useful data. Watching five hundred sites with the same constellation means accepting gaps. Priority tiering is an analytical and political decision that must be made before the architecture is fixed, not after.

What the customer owns and what the sensors cannot do

At handover, the operating state holds the satellites under its own ITU licence, the ground station hardware and software with full source-access terms, the processing and exploitation environment, trained national operators and analysts, and the complete archive of collected imagery and derived products. No imagery transits a Satellize or third-party server after handover; the chain of custody is national from the moment of downlink.

The limits are real and should be stated plainly. Optical resolution below roughly 30 cm requires very large apertures that push mass and cost beyond small-satellite economics; commercial systems at 30 to 50 cm resolution are sufficient to detect vehicles and construction activity but not to read markings or resolve small objects. SAR coherent change detection is sensitive to surface change but cannot determine intent; a disturbed patch of earth is a disturbed patch of earth until corroborated by other intelligence. Atmospheric gas detection is a screening tool, not a smoking gun: wind variability, sensor noise and source attribution uncertainty mean that a positive detection opens an investigation, it does not close one. None of these limitations are unique to a national programme; they apply equally to the systems used by the IAEA and the major intelligence agencies. The difference is that a national programme produces evidence the state controls.

Sovereignty terms and the single accountable engineer

Satellize structures verification programmes under the same sovereignty contract used across its space-stack: source-access terms agreed before signature, hardware audit rights at every integration milestone, and staged handover to national teams. The Satellize programme manager remains the single accountable engineer from mission architecture through to operational handover, which eliminates the coordination failures that arise when ground, space and software contracts are held separately.

After handover, Satellize retains no ongoing access to collected data. Optional long-term support arrangements cover hardware maintenance and software updates, but these are scoped separately and do not create dependency. The programme is designed to be operated by the national team without Satellize involvement. That is the point.

What this mission is built from

High-resolution panchromatic optical imagers: Provides sub-metre daytime imagery of declared and suspected sites, resolving facility configuration, vehicle presence and construction activity.
X-band SAR payloads: Delivers all-weather, day-night surface change detection using coherent change detection, identifying disturbed earth, new structures and displacement at centimetre scale.
Atmospheric sounders and gas spectrometers: Screens for anomalous gas concentrations, including SF6 and methane, near declared civilian or industrial sites as a probabilistic indicator for further investigation.
In-country data processing: Maintains an air-gapped, nationally controlled processing and exploitation environment that preserves chain-of-custody metadata required for evidentiary products.

What you end up owning

One or more satellites registered under the national ITU licence
Ground station hardware and software with full source-access terms and hardware audit rights
Air-gapped in-country processing and imagery exploitation environment
Complete downlinked data archive with provenance metadata intact
Trained national satellite operators and imagery analysts
Documented evidence-production workflow meeting the state's judicial or diplomatic standards

Handover proceeds in stages keyed to operator competency milestones: shadow operations, parallel operations and independent operations, typically spanning six to twelve months post-launch. At the final milestone, all administrative, operational and data-custody responsibilities transfer to the national team. Satellize retains no ongoing access to collected imagery or derived products; any subsequent support engagement is a separate, optional contract.

Programme parameters

Pathfinder configuration	1 to 2 satellites, optical or SAR payload, 1 ground station
Operational constellation	3 to 6 satellites, mixed optical and SAR payloads, 2 or more ground stations
Orbital regime	Sun-synchronous, 500 to 600 km altitude
Optical resolution class	30 to 50 cm panchromatic (small-satellite aperture constraints apply below 30 cm)
SAR change detection sensitivity	Centimetre-level surface displacement via coherent change detection, X-band
Revisit: pathfinder (1 satellite)	Approximately once per day at mid-latitudes for a given point target
Revisit: 4-satellite mixed constellation	Same-day revisit over priority sites; not continuous stare
Timeline to first satellite on-orbit	24 to 36 months from contract, dependent on payload heritage
National operator team at handover	Typically 8 to 15 trained personnel covering operations, ground and exploitation
Indicative cost class: pathfinder	Low tens of millions of dollars (consistent with publicly reported small-satellite missions of comparable payload class)

Sources

One contract, one accountable engineer

Commissioned as one programme, not a stack of contracts: spacecraft, launch, ground segment, mission control, training and handover are priced together. Source-access terms and audit rights are agreed in writing before signature. Request a verification architecture review.