Existing data to improve traffic information
With the number of vehicles, both private and commercial, in urban areas steadily increasing, accurate traffic information has become an increasingly important commodity. Conventional traffic-data collection systems have high infrastructure, maintenance and communication costs. However, with an increasing number of vehicles equipped with the Global Positioning System (GPS) there is also a large amount of useful data that is already potentially available. The ‘Creating a data mart for floating car data’ (TRACK AND TRADE) project looked to exploit this 'floating car data' (FCD) to solve a number of well-known problems that arise when generating traffic content. The EU-funded initiative looked to develop an FCD data mart that allows for the collection and integration of data from as many data suppliers as possible – using this 'dormant' resource as the basis for improved traffic-information services. This can then be exploited through a series of technological means. The main technological innovations include 'travel-time derivation technology'. This allows for simple collection of FCD from various resources, especially from sources such as taxi fleets, improving accuracy and running-time. The first step in the project was an extensive survey of existing data formats and standards in this field, from which the project team created an XML-based data model. They then developed a flexible data collection framework based on web services – simplifying the process of connecting new data sources to the data mart. A mediation framework allows for the mapping of any kinds of data source to their data model. The team created a software framework that uses a map-matching algorithm to relate the FCD to a road network and then calculate travel times. Because of the huge amounts of information involved, data management was crucial. The project defined a data warehouse schema, allowing for the aggregation of historic travel times to provide future estimates. Data structures were developed that are powerful enough to buffer live FCD data streams. Two graphic interfaces were created for different levels of users – a web-based solution based on open-source software and a more powerful proprietary solution. Exploitation of the results was a priority for the project partners, which included small and medium-sized enterprises (SMEs) and research institutes from several European countries. The partners implemented and tested five value-added services products and the intense collaboration has continued since the project ended.
