Through talking with customers and being personally involved in several projects which require having street data for routing purposes, I have learnt some valuable lessons which I want to share with the GIS community.
First of all, to get started working with street data is not as easy as it seems. One may need to contact a local agency for any street data in the public domain. At the same time, quite often there are no street data available, which implies spending quite a large amount of money on data acquisition. Technically, to get started performing network analysis using ESRI Network Analyst extension is relatively easy, yet to get the data for the work is much harder.
Secondly, there is a lack of awareness of street data sources available in the ESRI community. Obviously, there are much fewer people working with Network Analyst extension compared to the users of Spatial Analyst or 3D Analyst extensions. So, a person involved in solving routing tasks is forced to get to know the software on his/her own by using limited help system and other resources available. The number of white papers and magazine articles related to the ESRI tools for network analysis is limited as well, yet there is a great deal of interest in this particular area.
Finally, vendors’ street data is shipped in a raw format and obviously not in ArcGIS network dataset format! Navteq Navstreets and TeleAtlas Multinet products are delivered among others in ESRI Shape format. This implies that users may need to dive into the hundreds of pages which comprise the reference manual, in order to figure out many things – such as how to convert turns tables into ArcGIS turn feature class, import signposts, and set the directions parameters properly, to name a few. It is a time consuming process which is likely to involve contacting vendor’s support and extensive trial by error tests.
To give a thorough overview of the network analysis in ArcGIS would take too much time, so let’s just focus on the acquisition of street data in this post. Options available are as below/follows:
1) TIGER (Topologically Integrated Geographic Encoding and Referencing system) data from the U.S. Census Bureau is a good option to get the street data. Visit U.S. Census Bureau for more information or proceed to the ESRI ArcData Web interface for direct downloading available data. Street data do have a lot of attributive information, but one will still need to perform a lot of manual data editing. The data coverage includes only the USA.
2) ESRI has a separate product – ESRI StreetMap Premium, which provides access to the street data in SDC format both for North America and Europe. This product can be used in the Network Analyst directly and no data processing is required. Be aware that these network data are for read-only use and no edits to data can be applied. This data product provides the best performance as well as robust functionality, yet the least customization flexibility, because one can only use the street data as it is.
3) Navteq Navstreets and TeleAtlas Multinet are one of the most often used products for network analysis purposes. The data is delivered in MapInfo or ESRI Shape format and need to be processed in order to be used with the Network Analyst extension. Navteq does provide Navstreets data in the compiled ESRI File Geodatabase from the Q1 2011 yet some data processing still will be required. Depending on how much information you want to include into your network dataset, you will need to perform some joins and geometry processing operations as well as use ArcGIS geoprocessing tools or scripting language, for instance, Python. The data coverage includes the whole world.
4) OpenStreetMap – a collaborative Web project that aims at creating a free digital world map. One can download all the data including, but not limited to, street network data (http://downloads.cloudmade.com/). Street data do not have much attributive information so one would need to perform a lot of manual data editing. The data coverage includes the whole world.