Many public transport networks worldwide experience high crowding levels. Overcrowding can result in passengers not able to board the first arriving vehicle. We infer how waiting time induced by being denied boarding in crowded public transport systems is valued by passengers, based on observed passenger route choice behaviour. For this purpose, we estimate a revealed preference route choice model based on passenger and vehicle movement data. As denied boarding typically occurs only at specific locations and within strict time bands, whilst its occurrence is notoriously uncertain, we propose additional constraints to generate an appropriate choice set for which observed route choices can be used to estimate denied boarding perceptions. We found that the additional waiting time caused by denied boarding is valued 68% more negatively compared to the initial waiting time. On average, one minute of initial and denied boarding wait time are perceived as 1.62 and 2.72 min on-board an uncrowded vehicle, respectively. Not incorporating this more negative denied boarding wait time valuation can result in an underestimation of the passenger and societal impact of overcrowding in public transport systems. Moreover, it can underestimate the benefits of potential crowding relief measures and as such underestimate the benefit-cost ratio of these measures.@en

This paper synthesizes evidence from Workshop 8 ‘Big data in the digital age and how it can benefit public transport users’ of the 15th International Conference on Competition and Ownership in Land Passenger Transport. Big data in public transportation has increasingly attracted the attention from both scientists and practitioners, resulting in an increasing number of scientific studies and practical applications in this field. However, compared to the scientific developments, we see that practical big data applications are relatively limited, and that these are applied with a relatively low pace. This indicates that big data has not been used to its full potential in practice yet, meaning that public transport passengers currently do not fully benefit from the opportunities big data offers in terms of public transport quality and attractiveness. Based on literature study and input gained from a qualitative expert session with scientists, public transport authorities, public transport operators and transport consultants together during the conference workshop, we come to the conclusion that the challenges to stimulate further and faster use of big data in practice are institutional rather than technical. This complexity results from required coordination and cooperation among public and private entities that are not always aligned. A framework has been proposed with four components to stimulate a further and faster adoption of big data in practice, directing to different stakeholders or relations between stakeholders: align technical ambitions of big data applications with their institutional environment; enable/ease the use of big data by PT authorities by developing common definitions, data standards and consolidation; incorporate the use of big data by PT operators in the contract between authority and operator; quantify and visualize the business value of big data for PT operators. We illustrate our framework by successful case studies in Chile, the Netherlands and Sweden.@en

Robust public transport networks are important, since disruptions decrease the public transport accessibility of areas. Despite this importance, the full passenger impacts of public transport network vulnerability have not yet been considered in science and practice. We have developed a methodology to identify the most vulnerable links in the total, multi-level public transport network and to quantify the societal costs of link vulnerability for these identified links. Contrary to traditional single-level network approaches, we consider the integrated, total multi-level PT network in the identification and quantification of link vulnerability, including PT services on other network levels which remain available once a disturbance occurs. We also incorporate both exposure to large, non-recurrent disturbances and the impacts of these disturbances explicitly when identifying and quantifying link vulnerability. This results in complete and realistic insights into the negative accessibility impacts of disturbances. Our methodology is applied to a case study in the Netherlands, using a dataset containing 2.5 years of disturbance information. Our results show that especially crowded links of the light rail/metro network are vulnerable, due to the combination of relatively high disruption exposure and relatively high passenger flows. The proposed methodology allows quantification of robustness benefits of measures, in addition to the costs of these measures. Showing the value of robustness, our work can support and rationalize the decision-making process of public transport operators and authorities regarding the implementation of robustness measures.@en

In daily practice, public transport authorities and operators are constantly searching for improvements in public transport operations. To this end, it is necessary to identify inefficiencies and bottlenecks in the current public transport services. In this paper, we propose a method to automatically detect bottlenecks in the public transport network, using Automatic Vehicle Location data. A tool is developed to automatically process AVL data to identify bottlenecks for the current situation. This tool is applied to Amsterdam, capital of the Netherlands, where a new metro line will come into operation in the summer of 2018. The results show that bottlenecks are mainly found on radial lines and in the inner city. Therefore we expect that the operations of the tram network will improve in terms of operating speed and reliability due to the opening of the metro line, since the tram lines are expected to become less crowded and fewer lines will traverse the inner city.@en

More complex, efficient driver schedules reduce operator costs during undisrupted operations, but increase the disruption impact for passengers and operator once a disruption occurs. We develop an integrated framework to quantify the passenger and operator costs of disruptions explicitly as function of different driver schedule schemes. Since the trade-off between driver schedule efficiency and robustness can be quantified, this supports operators in their decision-making.@en

Overstappen heeft een grote invloed op de (on)betrouwbaarheid van het OV, een belangrijk kwaliteitsaspect voor veel reizigers. In eerdere studies wordt de impact van onbetrouwbaarheid van overstappen op de totale reistijd voor meerdere overstaprichtingen niet expliciet meegenomen. In deze studie presenteren we onderzoek naar synchronisatie van overstappen en de impact hiervan op betrouwbaarheid vanuit een reizigersperspectief. Reizigersbetrouwbaarheid is geanalyseerd voor een overstaplocatie tussen trein en tram, geëxploiteerd door twee verschillende vervoerders. We hebben een methode ontwikkeld om twee op de reiziger gerichte betrouwbaarheidsindicatoren te kwantificeren: extra reistijd ten opzichte van de dienstregeling, en benodigde buffertijd als maat voor spreiding. Door de beschikbaarheid van voertuiglocatie (NDOV) data en anonieme OV Chipkaartdata is het mogelijk om de betrouwbaarheid van een overstaplocatie, of van een OV netwerk als geheel, op netwerkniveau te kwantificeren. Vijf belangrijke variabelen zijn geïdentificeerd die betrouwbaarheid van een overstap kunnen beïnvloeden: de geplande overstaptijd, de verdeling van de werkelijke aankomst- en vertrektijden van de lijnen bij de overstaplocatie, frequentie van OV lijnen, overstap-looptijd, en het reizigersvolume. Voor zowel een hypothetisch OV netwerk, als voor een werkelijk OV netwerk, is aangetoond dat het verschuiven van de overstaptijd om één specifieke overstap te optimaliseren, effect heeft op de overstaptijd en betrouwbaarheid van andere overstappen. Deze methodiek geeft kwantitatief inzicht in deze trade-offs om uiteindelijk een optimum te vinden voor alle overstappen en reizigers samen. Dit kan worden meegenomen in maatschappelijke kosten-baten analyses. Met deze methodiek kunnen zowel de totale betrouwbaarheidsbaten, als de baten voor een specifieke groep reizigers, geïdentificeerd worden als gevolg van een bepaalde betrouwbaarheidsmaatregel.@en

A transit network design frequency setting model is proposed to cope with the postpandemic passenger demand. The multiobjective transit network design and frequency setting problem (TNDFSP) seeks to find optimal routes and their associated frequencies to operate public transport services in an urban area. The objective is to redesign the public transport network to minimize passenger costs without incurring massive changes to its former composition. The proposed TNDFSP model includes a route generation algorithm (RGA) that generates newlines in addition to the existing lines to serve the most demanding trips, and passenger assignment (PA) and frequency setting (FS) mixed-integer programming models that distribute the demand through the modified bus network and set the optimal number of buses for each line. Computational experiments were conducted on a test network and the network comprising the Royal Borough of Kensington and Chelsea in London.@en

Since ride-hailing has become an important travel alternative in many cities worldwide, a fervent debate is underway on whether it competes with or complements public transport services. We use Uber trip data in six cities in the United States and Europe to identify the most attractive public transport alternative for each ride. We then address the following questions: (i) How does ride-hailing travel time and cost compare to the fastest public transport alternative? (ii) What proportion of ride-hailing trips do not have a viable public transport alternative? (iii) How does ride-hailing change overall service accessibility? (iv) What is the relation between demand share and relative competition between the two alternatives? Our findings suggest that the dichotomy-competing with or complementing-is false. Though the vast majority of ride-hailing trips have a viable public transport alternative, between 20% and 40% of them have no viable public transport alternative. The increased service accessibility attributed to the inclusion of ride-hailing is greater in our US cities than in their European counterparts. Demand split is directly related to the relative competitiveness of travel times i.e. when public transport travel times are competitive ride-hailing demand share is low and vice-versa.@en

De last-mile in een openbaar vervoer (OV)reis is een van de meest hinderlijke gedeelten van een reis perOV,hierdoor is het OV veelal niet in staat om te kunnen concurreren met de auto. De oorzaak kan deels worden gevonden in het gebrek aan flexibiliteit en de lage snelheden die de veelal conventionele vervoersmiddelen op de last-mile kenmerken.Recente technologische ontwikkelingen makeninnovatieve vraaggestuurde vervoersconceptenmet zelfrijdende voertuigen mogelijk. Deze zijn onafhankelijk van vaste infrastructuur en zouden ideaalop de last-mile ingezetkunnen worden, om zo het diffusepatroon en lage vervoersvolume efficiënt tebedienen. Om de vervoerwaarde en de prestatie van een vraaggestuurd vervoersysteem met zelfrijdende voertuigen op de last-mile te bepalen is alscase:Station Delft-Zuid –TU Delft gekozen.Bij een vervoerwaarde studie metdergelijke innovatieve techniek wordt verwachtdat psychologische factoren, naast depuur instrumentele aspecten, expliciet een rol spelen in de vervoerwijzekeuzevan reizigers. Het meenemen van deze preferenties is daarom van belang, teneindede vervoerwaarde niet substantieel te over-of onderschatten. In dit paper zijnde uitkomsten van twee onderzoeken gecombineerd, een instrumenteel onderzoek (simulatiemodel + enquête) naar de vervoerwaarde van een systeem van zelfrijdende voertuigen en een stated preference (SP)-experiment om de invloed van diversepsychologische factoren op de vervoerwaarde te bepalen.De resultaten laten zien dat er een aanzienlijke vervoerwaarde bestaat voor zelfrijdende voertuigen op de last-mile, te weten57% van de steekproefpopulatie.Methetsimulatiemodel zijn diverse ITS maatregelengesimuleerd, deze maatregelen omvatten wijzigingen in de netwerkstructuur, voertuiggedragen het reizigersgedrag.De prestatie van het last-mile systeem, bleekveelalpositiefbeïnvloedt te worden door bovenstaande maatregelen.Effecten werden gemeten in een vergroting van de systeemcapaciteit, een reductie van de gemiddelde wachttijd ofwel van de gemiddelde reistijd. Het parallel toepassen van deze maatregelen realiseerteen aantrekkelijkere concurrentiepositie ten opzichte van de conventionele vervoerswijzen. De meest kenmerkende maatregel,het strategisch plaatsen van voertuigen,op locaties voorafgaand aan het ontstaan van vervoersvraag laat een reductie in de gemiddelde wachttijd van 40% zien.Uit het SP-experimentblijkt dat de attitudes ‘duurzaamheid’ en ‘vertrouwen’ de twee belangrijkste aspectenzijn in de vervoerwijzekeuze, terwijl de in-voertuigtijd niet minder negatief wordt ervaren dan in een door de reiziger gereden carsharing systeem. Dit suggereert dat de gebruikelijk genoemde voordelen van zelfrijdende voertuigenwellicht nog niet worden ervaren door de reiziger vanvandaag, en illustreert het belang om hier aandacht aan te schenken voor een succesvolle implementatievan zelfrijdende voertuigen.@en

In the recent years many developments took place regarding automated vehicles (AVs) technology. It is however unknown to which extent the share of the existing transport modes will change as result of AVs introduction as another public transport option. This study is the first where detailed traveller preferences for AVs are explored and compared to existing modes. Its main objective is to position AVs in the transportation market and understand the sensitivity of travellers towards some of their attributes, focusing particularly on the use of these vehicles as egress mode of train trips. Because fully-automated vehicles are not yet a reality and they entail a potentially high disruptive way on how we use automobiles today, we apply a stated preference experiment where the role of attitudes in perceiving the utility of AVs is particularly explored in addition to the classical instrumental variables and several socio-economic variables. The estimated discrete choice model shows that first class train travellers on average prefer the use of AVs as egress mode, compared to the use of bicycle or bus/tram/metro as egress. We therefore conclude that AVs as last mile transport between the train station and the final destination have most potential for first class train travellers. Results show that in-vehicle time in AVs is experienced more negatively than in-vehicle time in manually driven cars. This suggests that travellers do not perceive the theoretical advantage of being able to perform other tasks during the trip in an automated vehicle, at least not yet. Results also show that travellers’ attitudes regarding trust and sustainability of AVs are playing an important role in AVs attractiveness, which leads to uncertainty on how people will react when AVs are introduced in practice. We therefore state the importance of paying sufficient attention to these psychological factors, next to classic instrumental attributes like travel time and costs, before and during the implementation process of AVs as a public transport alternative. We recommend the extension of this research to revealed preference studies, thereby using the results of field studies.@en

It is important to understand how public transport passengers value on-board crowding since the outbreak of the COVID-19 pandemic. The main contribution of this study is to derive the crowding valuation of public transport passengers in a post-pandemic era entirely based on observed, actual passenger route choices. We derive passengers’ crowding valuation for the London metro network based on a revealed preference discrete choice model using maximum likelihood estimation. We find that after the passenger load on-board the metro reaches the seat capacity, the in-vehicle time valuation increases by 0.42 for each increase in the average number of standing passengers per square metre upon boarding. When comparing this result to a variety of crowding valuation studies conducted before the pandemic in London and elsewhere, we can conclude that public transport passengers value crowding more negatively since the pandemic. Furthermore, we found a ratio between out-of-vehicle time and in-vehicle time of 1.94 pre-pandemic and of 1.92 post-pandemic, based on which we conclude that the relative waiting/walking time valuation did not significantly change since the COVID-19 pandemic. Our study results contribute to a better understanding on how on-board crowding in urban public transport is perceived in a European context since the outbreak of the COVID-19 pandemic.@en

Urban metro and tram networks are regularly subject to planned disruptions, including closures, resulting from the need to maintain and renew infrastructure. In this study, we first empirically analyse the passenger demand response to planned public transport disruptions based on individual passenger travel behaviour, based on which we infer generalised journey time and cost elasticities for different passenger groups and time periods of the day. Second, we develop a model which enables predicting public transport demand for individual origin-destination pairs affected by a closure. The model is trained based on the empirically observed travel behaviour. The proposed method is applied to a case study closure in Amsterdam, the Netherlands, based on which we empirically derive generalised journey time and generalised journey cost elasticities of − 0.99 and − 1.11, respectively. Our results suggest that passengers’ demand response is lower for frequent users of the public transport network, as well as during weekdays - especially during the peak periods. Arguably, this stems from a higher share of captive passengers with a mandatory journey purpose in these segments, who will continue making their journey nevertheless. During weekends - with typically higher shares of leisure related journeys - a much more pronounced demand response is found. The estimated neural network regression model is able to predict passenger demand during public transport closures with a high level of accuracy. This provides public transport agencies more precise insights into the impact of closures on their revenue losses and on the potential need for resources reallocation.@en

Understanding the passenger demand impacts of public transport service changes is a fundamental aspect of transport planning. The main objective of this study is to derive an updated Generalised Journey Time (GJT) elasticity for urban and metropolitan public transport networks, by applying a revealed preference approach using individual passenger journey data. Based on more than 25 million empirical journeys subject to 9 different service interventions within the Greater London area, we find an average GJT elasticity of −0.61. The value implies that for every 1% increase in generalised journey time, on average public transport demand is expected to reduce by 0.61%, and vice versa. We also find that the demand response to service changes is most elastic during the midday period between the peak hours and most inelastic during the AM peak and early morning, possibly caused by a higher share of mandatory journeys. Our study results confirm the existence of a build-up rate from the initial short-run elasticity to a somewhat stronger longer-run elasticity. Besides, we find that at least within the short- and medium-term demand is more elastic to service degradations compared to service improvements. Our findings imply that it requires more time for demand to increase in response to a service quality improvement, compared to demand to decrease after a service quality reduction.@en

Disruptions in public transport can have major implications for passengers and service providers. Our study objective is to develop a generic approach to predict how often different disruption types occur at different stations of a public transport network, and to predict the impact related to these disruptions as measured in terms of passenger delays. We propose a supervised learning approach to perform these predictions, as this allows for predictions for individual stations for each time period, without the requirement of having sufficient empirical disruption observations available for each location and time period. This approach also enables a fast prediction of disruption impacts for a large number of disruption instances, hence addressing the computational challenges that rise when typical public transport assignment or simulation models would be used for real-world public transport networks. To improve transferability of our study results, we cluster stations based on their contribution to network vulnerability using unsupervised learning. This supports public transport agencies to apply the appropriate type of measure aimed to reduce disruptions or to mitigate disruption impacts for each station type. Applied to the Washington metro network, we predict a yearly passenger delay of 5.9 million hours for the total metro network. Based on the clustering, five different types of station are distinguished. Stations with high train frequencies and high passenger volumes located at central trunk sections of the network show to be most critical, along with start/terminal and transfer stations. Intermediate stations located at branches of a line are least critical.@en

Due to the multi-level nature of public transport networks, disruption impacts may spill-over beyond the primary effects occurring at the disrupted network level. During a public transport disruption, it is therefore important to quantify and control the disruption impacts for the total public transport network, instead of delimiting the analysis of their impacts to the public transport network level where this particular disruption occurs. We propose a modelling framework to quantify disruption impact propagation from the train network to the urban tram or bus network. This framework combines an optimisation-based train rescheduling model and a simulation-based dynamic public transport assignment model in an iterative procedure. The iterative process allows devising train schedules that take into account their impact on passenger flow re-distribution and related delays. Our study results in a framework which can improve public transport contingency plans on a strategic and tactical level in response to short- to medium-lasting public transport disruptions, by incorporating how the passenger impact of a train network disruption propagates to the urban network level. Furthermore, this framework allows for a more complete quantification of disruption costs, including their spilled-over impacts, retrospectively. We illustrate the successful implementation of our framework to a multi-level case study network in the Netherlands.@en

The availability of smart card data from public transport travelling the last decades allows analyzing current and predicting future public transport usage. Public transport models are commonly applied to predict ridership due to structural network changes, using a calibrated parameter set. Predicting the impact of planned disturbances, like temporary track closures, on public transport ridership is however an unexplored area. In the Netherlands, this area becomes increasingly important, given the many track closures operators are confronted with the last and upcoming years. We investigated the passenger impact of four planned disturbances on the public transport network of The Hague, the Netherlands, by comparing predicted and realized public transport ridership using smart card data. A three-step search procedure is applied to find a parameter set resulting in higher prediction accuracy. We found that in-vehicle time in rail-replacing bus services is perceived ≈1.1 times more negatively compared to in-vehicle time perception in the initial tram line. Waiting time for temporary rail-replacement bus services is found to be perceived ≈1.3 times higher, compared to waiting time perception for regular tram and bus services. Besides, passengers do not seem to perceive the theoretical benefit of the usually higher frequency of rail-replacement bus services compared to the frequency of the replaced tram line. For the different case studies, the new parameter set results in 3% up to 13% higher prediction accuracy compared to the default parameter set. It supports public transport operators to better predict the required supply of rail-replacement services and to predict the impact on their revenues.@en

Public transport disruptions can result in major impacts for passengers and operator. Our study objective is to predict disruption exposure at different stations, incorporating their location-specific characteristics. Based on a 13-month incident database for the Washington metro network, we successfully develop a supervised learning model to predict the expected number of disruptions, per type, station and time of day. This supports public transport authorities and operators to prioritize what type of disruptions at what location to focus on, to potentially achieve the largest reduction in disruption exposure. Our clustering results show that start/terminal and transfer stations are most susceptible to disruptions, mainly due to operations-and vehicle-related disruptions.@en

Disruptions in public transport have major impact on passengers and disproportional effects on passenger satisfaction. The availability of smart card data gives opportunities to better quantify disruption impacts on passengers’ experienced journey travel time and comfort. For this, accurate journey inference from raw transaction data is required. Several rule-based algorithms exist to infer whether a passenger alighting and subsequent boarding is categorized as transfer or final destination where an activity is performed. Although this logic can infer transfers during undisrupted public transport operations, these algorithms have limitations during disruptions: disruptions and subsequent operational rescheduling measures can force passengers to travel via routes which would be non-optimal or illogical during undisrupted operations. Therefore, applying existing algorithms can lead to biased journey inference and biased disruption impact quantification. We develop and apply a new transfer inference algorithm which infers journeys from raw smart card transactions in an accurate way during both disrupted and undisrupted operations. In this algorithm we incorporate the effects of denied boarding, transferring to a vehicle of the same line (due to operator rescheduling measures as short-turning), and the use of public transport services of another operator on another network level as intermediate journey stage during disruptions. This results in an algorithm with an improved transfer inference performance compared to existing algorithms.@en

Minimizing passenger transfer times through public transport (PT) transfer synchronization is important during tactical planning and real-time control. However, there are computational challenges for solving this Timetable Synchronization Problem (TSP) for large, real-world urban PT networks. Hence, in this study we propose a data-driven, passenger-oriented methodology as a preparatory selection stage to reduce problem dimensionality by (1) determining the significant transfer hubs in the network, and (2) identifying subsets of lines within these hubs that need to be prioritized for transfer synchronization. In the first phase of our methodology we determine the spatial boundaries of transfer locations, using a clustering technique based on the passenger transfer flow matrix inferred from smartcard data. After that, a subset of hubs to be prioritized for synchronization is selected. In the second phase, we characterize the transfer patterns within the hubs based on a topological representation. Based on these topological graphs, the line bundles that need to be prioritized within the hubs are further identified using a modularity-based community detection technique. We apply our methodology to a real-world case study, i.e. the PT network of The Hague, the Netherlands. For this case study, our approach allows for prioritizing 70% of all transfers within identified transfer locations while only requiring 0.9% of these transfer locations, thus reducing the complexity of solving the TSP substantially at a relatively low cost. Our method supports public transport operators during timetable design and real-time control in determining where and which lines to prioritize when devising measures for improving transfer experience and synchronization.@en