Bus travel time analysis is essential for transit operation planning. Then, this topic obtained large attention in transport engineering literature and several methods have been proposed for investigating its variability. Nowadays, the availability of large data quantities through automated monitoring allows more in-depth this phenomenon to be pointed out with new experimental evidence. The paper presents the results of some analyses carried out using automatic vehicle location (AVL) data of bus lines and automated vehicle counter (AVC) data on some corridors in the urban area of Rome where the bus services are mixed with other traffic and travel times are subject to high degrees of variability. The results show the effect of temporal dimension and similarity between travel time and traffic temporal patterns, and could open the road for the improvement of the short-term forecasting methods, too.

The opportunities offered by telematics development and the features required by travellers moving on more complex multimodal transit networks push to investigate new methods for generating path advice in transit trip planners. First, the study focuses on characteristics and limits of the methods used by current trip planners for path generation and then analyses the new methods applied by a new generation of trip planners (most at the prototypical developing stage). These methods use a group or, better, an individual traveller utility function, which allows personal preferences to be pointed out. As the individual utility functions in the transport modelling literature has been largely neglected, the second part of the study reports the state-of-the-art and some theoretical considerations on individual path utility function modelling, and recalls the approaches developed for including individual utility in new trip planners. Further, the results of individual utility function estimations, using surveys relative to the city of Rome, are presented and some aspects of the utility function calibration, including the info provider learning process of traveller preferences, are explored. The analyses show a significant improvement in using individual functions for path advice and a substantial influence of network complexity on the learning process.

This paper reviews the ex-post assessment of city logistics measures implemented in some European cities and, in a “what if” framework, proposes an analysis of tested environmental effects which may be useful in defining city logistics scenarios to be evaluated ex ante by simulation models. The analysis is performed in relation to the goals of environmental sustainability to pursue and the main characteristics of the cities in question (i.e., population and density). The paper aims to provide a tool that could be used in an ex-ante assessment methodology to identify a priori which measures (or set of measures) could best work in a specific city with respect to the environmental sustainability goals to pursue. Future scenarios can, thus, be readily defined and subsequently assessed by simulation tools in order to verify whether they meet the planned objectives. Although all measures can produce considerable environmental effects, the study shows that the choice of their implementation should be driven by the type of pollutant to detect.