The commercial speed of buses is a key factor in the operation of public transport systems, as it represents a direct measure of the quality of service provided to users as well as considerably affects the cost of the system. Monitoring the system performance by means of the commercial speed provided by the bus services is of course very important; however, in dense networks it becomes a very difficult task because of the amount of information required to implement such a monitoring procedure. The introduction of GPS technology in buses overcomes that difficulty in terms of information availability, although it brings along the challenge of processing huge amounts of information in a systematic way. We have been looking at the automatically collected data of the Santiago-Chile public transport system (Transantiago), where GPS observations of more than 6,000 buses operating on over 700 different routes are available every 30 seconds. In a one week period more than 40,000,000 positions of buses assigned to known routes are available. The commercial speed of buses can be estimated using this information. We developed a method to deploy in a synthetic and systematic way the average commercial speed of buses corresponding to each route, altogether as well as over segments of arbitrary length, during a whole day andalso divided in time intervals of arbitrary duration. This process can beconducted for a specific day and route or for any set of them. In this way, we transform a huge amount of data into useful information for the system planners and operators.
Abstract