Transit Services and User Information: an Application of Schedule-based Path Choice and Assignment Models

Transit Services and User Information: an Application of Schedule-based Path Choice and Assignment Models


U Crisalli, L Rosati, University of Rome Tor Vergata, IT


This paper applies a system supply, demand and assignment models for transit services defined through the innovative schedule-based approach. It can be used to assess the impact of the introduction of ATIS systems.


The use of innovative technologies applied to transportation systems allows to implement and use ATIS (Advanced Travellers Information Systems), aiming at improving individuals decision-making process (user?s optimum) through information to users on network performances in the attempt to facilitate their travel choices (e.g. departure time, route, etc.). ATIS system is usually implemented jointly with APTS (Advanced Public Transportation Systems), which aims at optimising performances at the overall system level (system optimum), for example for the efficient transit fleet management.
This paper applies a system supply, demand and assignment models defined through the innovative schedule-based approach to simulate current transit services, in order to give information about times and loads of transit vehicles according to the current transit system conditions, which can differ from the scheduled one due to reliability problems, and can influence user path choice.
In particular, concerning what (waiting times, on-board loads of upcoming buses, estimated travel time to destination, and so on) and where (pre-trip or en-route) user information is provided, it is possible to specify behaviourally consistent path choice models to simulate user choices in this ATIS environment, which allow us to calculate Level-of-Service attributes and on-board loads in a more precise way.
In addition to pathfinder services that users can use before starting their trip (pre-trip info), the latest ATIS system are able to provide ?customised? information to users (e.g. pay-per-info services) that via mobile phones can have pre-trip/en-route information on request regarding their own o/d trip.
The presented schedule-based models can be used to assess the impact of the introduction of ITS system, considering different levels of user info, which can be provided to all users (e.g., waiting times and loads of arriving buses at stop), or to specific classes (e.g., the estimated travel time to destination for each alternative, in addition to the previous info, provided for example via SMS). The latter information can affect to great extent en-route traveller?s choices, especially in congested transit systems, where travellers may choose to skip overloaded runs and wait for less crowded ones by trading off between longer waiting-time and higher on-board comfort.
The main characteristics of the schedule-based modelling framework are:
1)a time-varying O-D matrix estimation procedure based on real-time observation of number of passenger boarded and alighted from vehicles at stops;
2)a run-based (diachronic) supply model which allows to represent time-dependent transit services, whose temporal co-ordinates are updated real-time, in relation to the information on vehicle location provided by APTS;
3)a sequential time-dependent path choice model based on Random Utility Theory, which simulate user behaviour in relation to information provided by ATIS;
4)a within-day dynamic assignment procedure following a schedule-based approach, estimating the loads on each run of the transit system at any time of the reference period, which can be used to update user info, too.
An application example based on a real transit network (Ischia, Italy) will be presented, in order to verify the use of the proposed approach for operative planning.
In order to support the proposed paper according to the indication of the IM Committee:
The use of the schedule based approach in user choice path choice models specified according to user information (what, when and where it is given) is the innovative key of this work.
The modelling framework is self-contained and based on a strong theoretical basis, which has been developed and implemented to support operative planning.
Application examples are based on real networks, for which ATIS and APTS data where available for the scope of this research.
Public transport planners and operators, engineers and planners, as well as ; researchers and academics, can be interested in this research. In addition, please consider that only few transportation software on the market are able to consider this problem by the use of ?general-purpose? schedule-based models, which are not able to deepen transportation analysis and planning of transit services with ATIS systems.


Association for European Transport