Collecting transport infrastructure intelligence in Asia and the Pacific

July 2017

Current transport infrastructure data availability and quality often limit the development of sustainable transport policies, investment strategies and models of future transport needs.


Data gaps also constrain the ability to evaluate impacts of such policies and investments. Detailed and updated knowledge about the spatial distribution of transportation assets is relevant also to disaster risk reduction (DRR) policies. ADB is interested in understanding how satellite-based data collection could be included in a standard set of data collection methods for an institutional data collection mechanism. 

The EO support project supported the ADB Global Transport Intelligence – Transport Outlook Asia activity, a subproject of the Implementation of Sustainable Transport in Asia and the Pacific cluster CDTA (Capacity Development Technical Assistance). The CDTA targets three of the four key pillars of ADB’s Sustainable Transport Initiative Operational Plan. (STI-OP) The subproject in question (Subproject D) aims to help increase the knowledge base on transport in Asia and the Pacific and its impact on environmental and social sustainability. The CDTA project works with the developing member countries also to build local capacity for regular transport data collection and reporting. 

The satellite-based information contributed to the updated inventory of transport infrastructure (mainly roads, railroads, waterways, airports), and the identification of existing gaps, giving the most up-to-date picture of transport infrastructure, and of its evolution with respect to a previous date. Additionally, the activity also supported public transportation development projects both in the preliminary design and in the executive design phases, providing information supporting both the planning (e.g. population distribution around key points of interest such as train stations) and the detailed design activities (e.g. hyper-level of land use detail along a planned public transport corridor). The pilot areas of interest are Baku (Azerbaijan), Peshawar (Pakistan), Pohnpei (Federated States of Micronesia) and Suva (Fiji). A third service demonstrated the use of satellite-based information to deliver maritime traffic statistics based on integrating of coastal and satellite AIS (Automatic Identification System) data with detected vessels from radar EO data. The selected test areas for this service lie in the area from the Andaman to the Celebes seas.

Transport infrastructure inventory and its changes over time 

The knowledge of distribution and evolution over time of transport infrastructures is a key information for urban planning at local, regional and national level. The service demonstrating the detection of roads, railways, ports, airports and their associated areas, providing as much attributes as possible (e.g. surface type, number of lanes, construction status, etc.). The spatial resolution of the input satellite imagery proved crucial for the analysis level of detail. A higher resolution on the other hand lowers the possibility to cover large areas in a cost-effective way. To cope with such variability, imagery at different resolutions (from 0.5 m to 2.5 m) was tested on different. The tests showed that the highest resolution available is necessary to provide sufficiently accurate maps over relatively small urban areas, while for regional and national coverage a good compromise is to use 1.5 m resolution data. Naturally information could only be provided for above-ground assets and not always convincingly for the smallest elements such as footpaths and cycling lanes. 

Historical data availability allowed comparing the map of visible transport infrastructures around the year 2005 with the status around 2012. A third map was generated highlighting the changes that occurred in the period. This information if particularly efficient in reflecting the results of public investments in the transport network and is at the same time an indicator of the urban expansion in the area (particularly relevant in the Baku). Additional indicators can be derived about the connectivity of different urban zones in order to prioritise interventions. 

Detailed land use mapping and population estimation 

When planning new Mass Rapid Transit (MRT) systems or adapting existing ones, it is essential to have precise and up-to-date knowledge of the situation in the field. In this context, and following discussions with ADB CWUD, two types of products were defined and prototyped for greater Baku, where currently there are several options under evaluation to rearrange the railway line connecting Baku city and the surrounding urban areas. The products were:

  • Very detailed land use maps focussing on non-residential elements in a buffer area of 500 m width along the railway axis. This is basically an inventory of non-residential elements that may interfere with the planning of the adapted railway line. This product required a combined analysis of satellite images together with ancillary data from external sources to better characterise the use of the detected structures. The total length of the corridor was 80 km.

  • Population estimation within 1 km from the planned MRT stations.

This provided population information at the right level of detail and granularity, aggregated at building block level. Census data is usually outdated or collected in an aggregated way e.g. at district level. Its combination residential land use classes as detected form satellites makes it possible to estimate the population distribution at the level of the single building block. This information is relevant as it allows transportation planners to get reliable data on potential user basins. Population distribution (and density) maps were generated for 26 different stations. 

Maritime traffic statistics 

Radar satellites have the capability to detect vessels over large swaths of seas, both close to ports and at open sea. Vessels are detected thanks to their contrasting radar signature with respect to the sea background and can be characterised in terms of both size and speed. The acquired maritime situation can be completed with vessel identification data (e.g. AIS, SatAIS) in order to identify vessels not transmitting their identity and which, therefore, are suspicious. Radar satellites also offer the possibility to monitor large areas systematically (several times per day), allowing the generation of a vessel database and further statistics elaboration. Vessel identification information can be made available in a very short time frame (20 minutes after capturing the image, if a receiving station covering the area is available), thus allowing operational actions to be undertaken by national or international institutions. 

The service can be set up over any port area or to monitor specific areas (e.g. straits) or routes. 

Impact, benefits, sustainability 

Satellite-based products were demonstrated to be useful and effective in providing operational data to support the transportation planning process at different levels. The products are cost-effective when compared to other solutions for data collection and they guarantee systematic updates at the desired frequency (with a trade-off between resolution and frequent revisit) consistently over space and time.


The services delivered to ADB had beneficial effect on the planning of large area transport infrastructure inventories, which were identified as highly relevant for future creation of harmonised transportation databases across multiple bank member countries. The satellite services can follow first surveys of currently available transport infrastructure databases and may be used as a gap fuller solution. 

The detailed land use and population density products were extremely positive tive and ADB decided to further invest in the generation of such products to support other MRT feasibility studies in Peshawar and Karachi (Pakistan). One of the advantages of these products recognised by ADB staff is their independence of field surveys, especially in areas where field surveys may be difficult or impossible (e.g. for security reasons). Results from additional service provision over Karachi and Peshawar were considered positive, and steps are being taken to assess whether such products shall be considered as part of the standard methodology for new MRT feasibility studies. 

Project results and relevant EO capabilities were also presented at the ADB Knowledge Sharing Event How to Use Technology to Understand Urban Infrastructure and Human Mobility: A Hands-On Guide, jointly organised by ADB’s transport and urban sector groups at ADB Headquarters on 2015 September 17. There was general agreement that the introduction of geoinformation as a standard practice in the ADB urban and transport sector would be of great value, in particular for PPTAs (Project Preparatory Technical Assistance).

The benefits of using this technology for ADB is that we can plan and deliver our projects better, based on more reliable, better and more cost-effective information. For the same amount of money that we spent for planning previously, we can ask for more and better data, and deliver better services. That’s especially important for public transport planning, like for corridors of Bus Rapid Transit systems or even when looking at informal transport

Katja Schechtner,
ADB Transport Specialist

One of the major opportunities we had by using satellite imagery analysis is obtaining the data we needed without going into the field. Satellite imagery allowed us to scan completely the corridor where we work, and to get all the data for the utilities, for the width of the corridor. Everything present in the corridor from one facade to the other was perfectly represented in the satellite imagery. It also allowed us to analyse the land use around the mass transit stations that we are planning. The outcome of the collaboration was very good.

David Margonsztern,

ADB Senior Urban Development Specialist (Transport)

  • Livablecities
  • Livablecities
  • LinkedIn Social Icon