4.3.2 — Smart Ports — maturity: live
Container Yard Monitoring
Using very-high-resolution optical satellite imagery and change-detection algorithms to continuously audit container stack density, dwell time, and yard utilisation across national port terminals.
Satellite imagery and AIS fusion give port authorities a live, vendor-independent picture of every container stack, gate queue, and yard movement — no ground sensors required.
Port authorities and customs agencies rarely have a real-time, independent picture of what is actually sitting in their container yards. Ground-level CCTV covers lanes, not the full yard geometry; terminal operating systems reflect what was declared, not what is physically present. The gap between manifest data and ground truth is where smuggling, mis-declared cargo, and customs revenue leakage hide.
A constellation of very-high-resolution optical microsatellites, tasked on a daily or sub-daily schedule, produces orthophotos of every major national terminal at 30–50 cm resolution. Computer vision pipelines count individual TEUs, classify stack configurations, flag anomalous dwell times, and detect overnight movements that bypass formal gate processes. Cross-referencing satellite-derived container counts against customs manifests and TOS exports exposes discrepancies that ground inspection can then target.
The operational outcome is twofold. Customs and border agencies gain an independent verification layer that is impossible to tamper with at the terminal level, because the sensor is overhead and sovereign. Port planners gain objective utilisation metrics across all terminals simultaneously — not self-reported figures from competing terminal operators — enabling smarter infrastructure investment and accurate throughput forecasting without relying on data that commercial operators have every incentive to massage.
Frequently asked
Can satellites actually count individual containers in a busy yard?
Yes, at sub-0.5 m resolution — commercially available from operators such as ICEYE and Capella — individual 20- and 40-foot containers are distinguishable in both optical and SAR imagery. Machine-learning object-detection models routinely achieve container-count accuracy above 90% under clear-sky or all-weather SAR conditions. Occlusion from cranes and stacked rows above four high reduces accuracy in the densest blocks.
Why should a government own the satellites rather than subscribe to Planet or Capella?
Commercial tasking contracts can be suspended, reprioritised, or priced out of reach during a crisis — exactly when port intelligence matters most. A sovereign constellation guarantees tasking priority, keeps raw imagery under national jurisdiction, and allows the government to share data freely with customs, coast guard, and logistics agencies without licensing restrictions. The capital cost is recovered through avoided dwell-time losses and reduced demurrage.
What orbit and satellite class makes sense for port monitoring?
A LEO constellation of 6–12 microsatellites (50–150 kg) in a 450–550 km sun-synchronous orbit provides the best trade-off between ground resolution, revisit frequency, and launch cost. Sun-synchronous passes deliver consistent solar illumination for optical sensors; SAR payloads on the same bus operate day-night and through cloud. Nanosatellites below 10 kg generally cannot carry the aperture needed for sub-1 m resolution.
How does AIS data complement satellite imagery for yard monitoring?
AIS identifies which vessel is berthed or at anchor, its cargo manifest linkage, and its ETA/ETD — context that a pixel-counting algorithm cannot derive from imagery alone. Fusing S-AIS (space-based AIS, collected by the same LEO satellites) with yard imagery lets analysts trace a container's journey from ship discharge to gate-out. Spire and Unseenlabs provide S-AIS as a payload-of-opportunity on small satellites, making combined missions cost-efficient.
What is the realistic end-to-end latency from satellite capture to actionable alert?
With direct downlink to a national ground station and automated cloud processing, latency from image capture to georeferenced container-count delta is typically 15–45 minutes. Adding an AI triage layer for anomaly flagging (e.g. unauthorised container movement at night) adds 2–5 minutes of compute time. This is adequate for port scheduling and customs prioritisation, though not for real-time crane control.
Does this work for inland container depots as well as sea ports?
Yes. The same VHR optical and SAR techniques apply to any paved yard with stacked ISO containers, including inland container depots (ICDs) and dry ports. The World Bank's 2023 port-reform guidance explicitly recommends satellite monitoring for dry-port capacity planning in landlocked developing countries.
How does weather affect SAR versus optical for this use case?
Synthetic Aperture Radar penetrates cloud, rain, and darkness, making it the operationally reliable choice for ports in tropical or high-latitude regions. The trade-off is that SAR imagery requires more specialist interpretation: metal containers produce strong returns but shadow effects from adjacent stacks can suppress detections. Fusing SAR and optical when both are available improves detection rates significantly.
What cybersecurity obligations apply to a government satellite ground system feeding port data?
IMO MSC.428(98) requires flag states and port operators to address cyber risk in their Safety Management Systems by 2021 (now in force). A national ground-station-to-port-authority data link must meet these obligations plus any national critical-infrastructure protection rules. End-to-end encryption, authenticated command uplinks (per CCSDS 352.0-B-2), and segregated operational networks are the baseline architecture.