Detecting Trucks and Parking Availability with Satellites and Drones: SPATRA’s Road Pilot

SPATRA webinar on Rail Workstream

Efficient cross-border transport depends on reliable traffic flow and access to safe parking. Congestion at international borders and uncertainty about parking availability cause delays, additional fuel use, and safety concerns. SPATRA addresses these issues in its road pilot by developing a monitoring approach that combines Earth observation, in-fleet data, and drone imagery. The aim is to improve estimated arrival times (ETA), planning of mandatory driver breaks, and logistics efficiency.

 

The challenge of satellite truck detection

Detecting heavy vehicles from space is not straightforward. Sentinel-2, part of the Copernicus programme, provides 10-metre resolution optical images. A truck, however, is smaller than a single pixel and cannot be directly observed. To overcome this, SPATRA applies discrete remote sensing models and machine-learning methods trained on labelled targets. These algorithms can identify signatures associated with trucks even when they are below pixel size.
This makes it possible to estimate the density and movement of heavy vehicles across border-crossing areas and selected parking facilities using Sentinel-2 imagery.

 

Extending coverage with drones and EGNSS


Because Sentinel-2 has a revisit cycle of around five days, it cannot provide near real-time congestion data. To complement this, SPATRA integrates EGNSS-enabled drones equipped with cameras:

  • EGNSS positioning ensures accurate flight paths and image geo-referencing.
  • Onboard cameras capture high-resolution images that are used to train and validate detection algorithms.
  • Drones supply a current view of border queues and parking occupancy when deployed, supporting operational decisions.

This hybrid approach offers both broad, historical insights from satellites and up-to-date local observations from UAVs.

Parking occupancy snapshots
Planning driver breaks is mandatory under EU rules, but parking availability is often uncertain. In SPATRA, drone imagery is used to provide snapshots of parking occupancy, showing the situation at the time of flight. This visual evidence complements satellite and in-fleet data and helps operators decide where breaks can be taken more efficiently. The system does not forecast occupancy; it provides verified observations when UAVs are deployed.

Integration with logistics operations
The road pilot is carried out with Nelt, a logistics operator in Southeast Europe. Nelt supplies in-fleet data from EGNSS-enabled trucks, enhancing accuracy. By combining EO, UAV, and in-fleet data, the project aims to provide:

  • More accurate ETAs for border crossings
  • Alerts on deviations from planned routes
  • Practical information to support planning of rest breaks

Expected outcomes
SPATRA has set clear key performance indicators for the road pilot:

  • At least 90% accuracy in arrival time predictions
  • Over 85% accuracy in detecting deviations from planned routes
  • Demonstrated reductions in congestion-related CO₂ emissions through optimised routing

This pilot demonstrates how combining Copernicus Earth observation, EGNSS, and drone technologies can deliver actionable insights for the transport sector. By detecting trucks and providing reliable snapshots of parking occupancy, SPATRA supports safer, more efficient cross-border freight operations and contributes to reducing environmental impact.