ESA title
Library / EO Capabilities / Transport Network

Transport Network

Add to E-Portfolio ADD TO E-PORTFOLIO
Report an issue SUGGEST AN EDIT
Transport & InfrastructureUrban Sustainability Operational Use
Add to E-Portfolio ADD TO E-PORTFOLIO
Report an issue SUGGEST AN EDIT

EO Capability Benefits

Accurate information on the location, type and condition of transport links is essential for network analysis and infrastructure planning. Access to this type of data allows planners to map differences in connectivity across an Area of Interest, especially in towns and cities, and understand where investments in extensions or improvements will create the greatest impact. Transport network data is also essential for transport modeling to identify existing bottlenecks and simulate the effect of network changes on overall transport efficiency and the cumulative decisions of network users. Finally, these datasets are also a vital component of disaster risk management as planners can simulate how road closures, for instance due to flooding, impact the network as a whole as well as efficiency of emergency response.

EO Capability Description

Geospatial datasets depicting an inventory of the Transport Network are provided in vector format, consisting of segments or links connected via nodes. A wide range of attributes can be attached to each segment indicating, for instance, its name, type (road, rail, waterway), place in the road hierarchy (e.g. arterial, collector, local), width, type of surface, speed limit, access restrictions or direction of travel. Metadata such as this can be used by traffic modelling software and GIS applications, e.g. for creating more realistic isochrones (areas accessible from a certain starting point in a certain period of time using a chosen mode of transport).

Using deep learning techniques, commercial sub-meter Very High Resolution (VHR) imagery can be used to automatically detect road surfaces and thus provide a foundation for subsequent, sometimes manual, error correction as well as labeling and adding of attributes. Recently, promising results have also been achieved with super-resolution Sentinel-2 data – that is high resolution non-commercial imagery whose spatial resolution has been enhanced (“upsampled”) using algorithms trained for this purpose.

Discover this EO Capability on the Analytics & Processing Platform (APP)

AI Road Extraction
Indicative Cost

Free

OpenStreetMap data is often sufficient for most use cases relying on a vector format transport network in many (but not all) regions. Show more

Relevant EO Technologies
VHR OPTICAL

Very High Resolution (VHR) optical imagers are passive, nadir‑viewing radiometers that measure reflected solar radiation in a limited set of broad spectral bands, using pushbroom or similar designs to build 2‑D images as the satellite moves along its orbit. Depending on definition, typical spatial resolutions for these systems range from about 5m to less 1m.

Related Training Resources

APP links