Find in Library

Road-based mass transit systems are an effective means to combat the negative impact of transport that is based on private vehicles. Providing quality of service in this type of transit system is a priority for transport authorities. In these systems, travel time (TT) is a basic factor in quality of service. This paper presents a methodology, based on data mining, for analyzing TT in a mass transit system that is planned by timetable. The objective of the methodology is to understand the behavior patterns of TTs on the different routes of the transport network, as well as the factors that influence these patterns. To achieve this objective, the methodology uses clustering techniques to process the GPS data provided by the vehicles of the public transport fleet. The results that were obtained when implementing this methodology in a public transport company are presented as a use case, demonstrating its validity.