Investigating bus travel time and predictive models: a time series-based approach

Public transport agencies observe the travel time as one of the main parameters of urban transport performance. In particular, travel time forecasting is an important planning tool for public transport companies given that it can improve the quality of the planned services by reducing the gap between the actual and the planned travel times. In this paper, this relevance is discussed and based on the experimental evidences the goodness to use time series based approach is pointed out. In fact, among the large number of factors affecting the operation of public transport, most of them are shown to follow a given temporal pattern. The analysis is performed using data from automated vehicle monitoring of buses lines sharing the way with other traffic in Lviv (Ukraine). The results prove the goodness of such an approaches and the opportunity offered to operators to improve their services.