Overcrowding is a major phenomenon affecting travel experience in urban public transport, whose negative impacts can be potentially mitigated with real-time crowding information (RTCI) on public transport vehicle departures. In this study, we investigate the willingness to wait (WTW) with instantaneous RTCI to avoid the in-vehicle (over)crowding the passenger faces, focusing specifically on urban crowding context (i.e. bus and tram systems). We conduct a stated-preference survey in Krakow (Poland), where we examine the choice probability between boarding now a more crowded vehicle vs. waiting at the stop for a less-crowded PT departure, and estimate a series of discrete choice models. Results show that 50–70% of respondents consider skipping a first departure which is excessively overcrowded and 10–30% would skip a vehicle with moderate standing crowding on-board. Acceptable waiting times typically range between 2 and 13 min, depending on crowding level and propensity to arrive on-time, but may even exceed 20 min in individual cases. These findings indicate that RTCI can induce a substantial WTW, affecting travel behaviour. We discuss its implications for mitigating service disruptions and demand management policies, including prospective support for public transport recovery in the aftermath of covid-19 crisis.

We propose a method to evaluate public transport network vulnerability. We study the evolution of the passenger Volume Over Capacity (VOC) ratio throughout the network to measure the spatial and temporal extent of the impacts caused by an unplanned service segment disruption. The VOC ratio provides an indication of the on-board travel comfort, an important level-of-service indicator, as well as reflects the residual capacity for absorbing additional demand. Because of the dynamic nature of public transport systems, disturbances propagate through the network in both time and space. Our modelling approach is able to capture transit system dynamics and quantify the extent to which the network exhibits spillover effects. We apply the method to the case of the rapid public transport system of Stockholm Sweden We demonstrate how the changes in network saturation and the corresponding recovery time can be quantified.

The objective of this paper is to understand the consequences of providing real-time information on crowding levels (RTI-CL) in public transport networks. We propose to extend the mesoscopic, simulation-based assignment model with the passengers' knowledge of instantaneous crowding levels onboard the public transport vehicles. We illustrate the results on a sample transit network, where we investigate the arising changes in network performance and journey experience as a result of RTI-CL provision. We demonstrate that effects of providing the crowding information in real-time on en-route path choices is strongly related to number of assumptions that we address in the paper. The resulting changes are dependent on: network congestion level (the effects are mostly pronounced in moderately congested networks), passengers' behaviour (increased sensitivity to crowding induces higher variability in vehicle loads), the RTI-CL penetration rate (network performance becomes worse with ubiquitous access to information) and information provision type (smoothed information over recent vehicle runs leads to higher accuracy than instantaneous information based on the single latest run only). Based on these, we formulate conclusions for further studies, larger-scale applications and practical implementation of real-time passenger crowding information systems.