Civil unrest moves fast and rarely announces itself. A protest that fills a square at noon can become a critical infrastructure threat by dusk, and ground-level intelligence is routinely denied or degraded precisely when commanders need it most. Conventional aerial surveillance requires pre-positioning assets and carries its own escalation risk; renting commercial tasking from a foreign vendor means a third party decides what gets imaged, when, and whether your government receives it at all.
A sovereign constellation closes that gap. Sub-metre optical revisits combined with night-capable SAR allow analysts to track crowd density, movement corridors, vehicle concentrations and perimeter breaches across multiple simultaneous flashpoints without exposing personnel. RF survey payloads map mobile communications activity — a reliable proxy for crowd size and coordination tempo — and flag anomalous emission patterns that precede organised violence.
The operational output is a live common operating picture for national police, gendarmerie or emergency management commands: crowd heat-maps updated every 30–90 minutes, automated alerts when density thresholds or vehicle counts exceed doctrine-defined triggers, and a time-stamped archive that is immediately available as legal evidence. Owning the architecture means the intelligence flow is never suspended by a vendor's political calculation or a UN Security Council disagreement.
Frequently asked
Why can't we just task a commercial constellation like Planet or BlackSky during an emergency?
You can — until you can't. Commercial operators are incorporated under foreign jurisdictions, subject to their governments' shutter-control orders and export licence conditions. During a geopolitically sensitive domestic event, a foreign government can legally compel a commercial operator to deny tasking or embargo imagery delivery. A sovereign constellation has no such chokepoint. Owning the asset means owning the decision.
What resolution do we actually need to monitor a crowd or protest site?
Academic research and operational experience converge on 0.5–1 m GSD as the practical threshold for crowd density estimation, vehicle identification, and perimeter assessment. Sub-0.5 m is useful for individual-level detail but triggers significantly heavier legal scrutiny around privacy. A well-calibrated 0.5 m microsatellite constellation balances operational utility against proportionality requirements under most human rights frameworks.
How does SAR complement optical imagery for this application?
Synthetic Aperture Radar penetrates cloud, smoke, and darkness — the three conditions most likely during a serious civil disorder event (tear gas, fire, night operations). SAR cannot distinguish crowd density as cleanly as optical at equivalent resolution, but it reliably detects large vehicle concentrations, changed surface conditions, and structural damage. The architecture of choice pairs an optical cluster for day/clear conditions with a SAR payload for persistent all-weather awareness.
Is this use case legal under international human rights law?
Overhead surveillance of public gatherings is not per se unlawful, but it must satisfy the proportionality, necessity, and legality tests under instruments such as the UN International Covenant on Civil and Political Rights (ICCPR Article 17) and regional equivalents. Nations should establish a statutory authorisation regime specifying the threshold of disorder that triggers satellite tasking, maximum data retention periods, independent oversight, and redress mechanisms before operational deployment.
How many satellites do we need for a minimum viable constellation?
A three-satellite LEO optical constellation in a single orbital plane provides daily revisit for most latitudes — sufficient for post-event damage assessment and pattern-of-life analysis, but inadequate for near-real-time monitoring. Operational utility for reactive public safety intelligence starts at six satellites across two orbital planes (~45-minute revisit) and scales toward 12–18 for sub-15-minute coverage. Cubesat or microsatellite buses in the 50–150 kg class are now the standard architecture for such constellations.
Can AI automatically detect and classify unrest events from imagery?
Computer vision models trained on labelled overhead imagery can flag anomalous crowd concentrations, unusual vehicle movements, and infrastructure blockages with useful accuracy — typically 75–88% precision on benchmark datasets. However, false-positive rates remain operationally significant, and the models degrade under novel event types or unfamiliar urban morphologies. Current best practice treats AI output as a cueing and triage layer, with human analysts making consequential decisions.
What happens to the imagery data after an event — who controls it?
Under a sovereign-owned model, the nation's own data governance law governs retention, access, and deletion. Under a commercial purchase arrangement, the vendor's terms of service, their government's surveillance laws, and any applicable export control regimes all apply — potentially granting foreign actors lawful access to sensitive operational imagery. This is a non-trivial national security consideration that data sovereignty advocates at UN-OOSA consistently flag as underappreciated.
What ground infrastructure does a sovereign riot-monitoring constellation require?
At minimum: one primary satellite operations centre with mission planning and telemetry capability, at least two geographically diverse ground stations for downlink resilience, a secure image processing and exploitation pipeline, and analyst workstations integrated with GIS tooling (e.g., OGC-compliant platforms). Operational hardening — including redundant power, hardened communications, and access controls meeting NIST SP 800-53 — should be considered from day one rather than retrofitted.