Posts tagged APC

De halteparadox – een balans tussen meer en minder haltes

Afgelopen jaar was er veel aandacht voor een bekend dilemma in het ontwerp van ov-lijnen: hoeveel haltes zorgt voor het optimum voor reizigers? Tim de Ridder, student aan de TU Delft, zocht voor de Haagse vervoerder HTM naar dit optimum. “Met minder haltes kan HTM een betere kwaliteit bieden”.

Lees het hele artikel uit OV Magazine HIER

Het onderzoeksrapport met gedetailleerde resultaten is HIER te vinden

Gerelateerd artikel: “Minder bushaltes alléén niet gelijk aan minder bereikbaarheid”


Podcast Mobility Innovators: Human-centered design for Smart Public Transport

Technology and New mobility are reshaping urban transportation in cities. Human-centric design is key to the quality of life in cities, putting people at the heart of urban transport planning. All stakeholders, including academia, will play a key role to reshape the future of mobility.

Listen to the podcast of Mobility Innovators with Niels van Oort:

04:00 Service reliability in public transport

07:40 About Smart Public Transport Lab at Delft University

14:00 How to run LRT system in the cities efficiently

20:20 Digital Inequality in Transport Services

28:50 Tesla predication on Self Driving Vehicles

34:50 MaaS from the passengers’ perspective

38:30 First & Last miles connectivity

44:54 Use of Big Data to improve services

49:05 Role of academia in the new world

Find more details about the discussed topics here:

Digital inequality (literature review paper)

Service reliability (podcast and papers)

5E model of wider impacts of public transport (book chapter 6, page 112-)

Light rail, lessons from 61 cities (book)

Bicycle+transit combination (podcast+papers)

Amsterdam North/South metro line impact study

The four-year study conducted by TU Delft, AMS Institute and others into the transport-related, spatial and economic effects of the North/South metro line is now complete and was presented to Amsterdam’s regional and City councils end of 2021.

Find the final policy report here HERE

Find the main TU Delft research findings HERE

Find the interactive visualisations of the GVB timetable and anonymous passenger data before and after HERE

Find all research papers and reports HERE

MSc thesis Simon van Hees: Regional Travel Time and Transfer Impacts of the Noord/Zuidlijn using Interoperable Smart Card Data

Impact assessment of new North/South metro line in Amsterdam

Large infrastructural projects are usually evaluated ex-ante before the decision to build the project is taken. However, after construction and opening of such project a thorough ex-post analysis is rare. In this paper we present an overview of such an evaluation study conducted in Amsterdam, capital of The Netherlands, including some first results. Research themes in the study are public transport, mobility and accessibility, public space and liveability and spatial economics. In this paper we focus on effects on public transport.

The new north-south metro line in Amsterdam became operational in summer 2018. This was accompanied by changes to the existing bus and tram network to provide feeder services to the new line, as well as to remove duplicate routes. Apart from adding significant capacity to the public transport network, the new line and the accompanying changes to the network are expected to improve travel times, reliability, accessibility and comfort levels (at least on average; not for all individual travellers).
The changes in such service quality attributes is expected to lead to a change in travel behaviour in terms of public transport route choice, mode choice (between public transport and private modes or within public transport), destination choice, departure time choice or addition of new trips (induced demand).
The objective of this study is to identify the main effects of the new metro line on existing and new passengers. We pay attention to the following aspects:
– Passenger volumes.
– Travel times, where the following distinction can be made:
o in-vehicle time;
o waiting time at the first stop;
o transfer walking time;
o transfer waiting time.
– Number of transfers.
– Network flows / crowding in vehicles.
– Reliability: travel time variance on the journey level.
– Accessibility: number of inhabitants and jobs reachable within a travel time budget.

Data sources for the study are GTFS timetable data (open source), Smart card data (both from within the city of Amsterdam as for the regional feeder bus services) and Automated Vehicle Location data. To measure perceived quality of the PT network, a survey is conducted among inhabitants of Amsterdam. In this survey approximately 3.800 respondents were asked about the travel time perceptions of their last PT trip, both before and after opening of the metro line. Finally, for a sample of travellers the entire trip is followed by a GPS tracking app.

Impact analysis of a new metro line in Amsterdam using automated data sources

A new metro line (the north-south line) was opened in Amsterdam in July 2018, adding significant capacity to the existing urban public transport network consisting of bus, tram and metro modes. The opening of the metro line was accompanied by changes to the existing bus and tram network, such as removal of duplicate routes and addition of feeder routes.

Traditionally, the impact of such a network change was measured either ex-ante or post-op based on surveys or model forecasts (Vuk 2005; Knowles 1996; Engebretsen, Christiansen, and Strand 2017). However, with the availability of automated data sources such as the smart card data, the exact impact on transit demand and service quality can now be measured. However, so far this has been limited to analysing the changes in travel times and reliability at a trip level (Fu and Gu 2018), excluding transfers.
This research utilises smart card and AVL data to study the impact of the new line on travel patterns (passenger flows), travel times and reliability from a passenger perspective by considering journeys including transfers. The metrics are calculated at a stop-cluster level, enabling also a distributional analysis of the impacts. Such a post-op analysis of any policy intervention or network change could be used to refine the demand predictive ex-ante tools.

Check the Transit Data workshop contributions of Malvika Dixit: Presentation and Extended abstract

Forecasting bus ridership with trip planner usage data: a machine learning application

Currently, public transport gives much attention to environmental impact, costs and traveler satisfaction. Good short-term demand forecasting models can help improve these performance indicators. It can help prevent denied boarding and overcrowding in busses by detecting insufficient capacity beforehand. It could be used to operate more economically by decreasing the frequency or the size of the bus if there is overcapacity. Moreover, it could help operators plan their busses during incidental occasions like big public events where little information is known. Finally, it could be used to reliably inform the travelers on the current crowdedness.
This study investigates the usefulness of a new data source; the usage data of a trip planner. In the Netherlands there are multiple trip planners available for users to help find the most optimal (multimodal) journeys. These trip planners require a date, a time and an origin and destination, which they use to construct multiple alternative journeys from which the user can choose. For this study the data of 9292 was used, being the major trip planner in the Netherlands including all public transport modes.
We developed a model for forecasting the number of people boarding and a model for forecasting the number of people alighting at a certain stop. These forecasts are defined at the vehicle-stop level. By summing the number of people boarding and subtracting the number of people alighting along the trip the forecasted number of passengers after a stop is calculated.

We compare five different machine learning models: multiple linear regression, decision tree, random forests, neural networks and support vector regression with a radial basis kernel. We compare these models with two simple rules: 1 predict the same number as last week, and 2 predict the historic average as number. The models are implemented in the Scikit-Learn library of Python. The data is stored in a PostgresSQL database.
The trip planner datasets and smart card dataset are merged and preprocessed. The resulted dataset is rather sparse; a lot of stops have zero passengers boarding or alighting or requests suggesting to do so. Therefore we investigated if subsampling is needed. From the datasets useful data is selected and features are constructed. The features are standardized. Different number of features are tested, these features are selected based on recursive elimination using a simple random forests model. Finally, the hyperparameters of the models are tuned and the optimal configurations are stored. The scores are validated by using cross validation.

Find more details in the following contributions by Jop van Roosmalen: Transit Data workshop presentation and MSc thesis

Combining Speed Adjustment and Holding Control for Regularity based Transit Operations

Vehicle bunching often occurs in high-frequency transit systems leading to deterioration of service reliability. It is thus necessary to control vehicles during operations. Holding control is a common solution for this situation, but it may result in longer vehicle running times. Speed adjustments can contribute to more regular operations while preventing prolonged trip times. This paper proposes a control strategy, which combines these two strategies to maintain the regularity of transit operations. The findings based on simulation study for trunk bus services in the Netherlands suggest that combining the two strategies implies both the positive and negative attributes of each control.

Find the MT-ITS presentation and paper of Aishah Imram HERE and HERE

Passenger Travel Time Reliability for Multi-Modal Public Transport Journeys

Urban transit networks typically consist of multiple modes and the journeys may involve a transfer within or across modes. Hence, the passenger experience of travel time reliability is based on the whole journey experience including the transfers. Although the impact of transfers on reliability has been highlighted in the literature, the existing indicators either focus on uni-modal transfers only or fail to include all components of travel time in reliability measurement. This study extends the existing ‘Reliability Buffer Time’ metric to journeys with multi-modal transfers and develops a methodology to calculate it using a combination of smartcard and automatic vehicle location data. The developed methodology is applied to a real-life case study for the Amsterdam transit network consisting of bus, metro and tram services. By using a consistent method for all journeys in the network, reliability can be compared between different modes or between multiple routes for the same origin-destination pair. The developed metric can be used to study the reliability impacts of policies affecting multiple modes. It can also be used as an input to behavioral models such as mode, route or departure time choice models.

Find the TRB paper and presentation of Malvika Dixit HERE and HERE

Robust Control for Regulating Frequent Bus Services: Supporting the Implementation of Headway-based Holding Strategies

Reliability is a key determinant of the quality of a transit service. Control is needed in order to deal with the stochastic nature of high-frequency bus services and to improve service reliability. In this study, we focus on holding control, both schedule- and headway-based strategies. An assessment framework is developed to systematically assess the effect of different strategies on passengers, the operator and transport authority. This framework can be applied by operators and authorities in order to determine what holding strategy is most beneficial to regulate headways, and thus solve related problems. In this research knowledge is gained about what service characteristics affect the performance of holding strategies and the robustness of these strategies in disrupted situations, by using scenarios. The framework is applied to a case study of a high-frequency regional bus line in the Netherlands. Based on the simulation results, we identified the line characteristics that are important for the performance of schedule- and headway-based strategies and determined how robust different strategies are in case of disruptions. Headway-based control strategies better mitigate irregularity along the line, especially when there are disruptions. However, schedule-based control strategies are currently easier to implement, because it does not require large changes in practice, and the performance of both strategies is generally equal in regular, undisrupted situations. In this paper, insights into what the concerns are for operators with respect to technical adaptations, logistical changes and behavioral aspects when using a headway-based strategy are given.

Find the TRB paper and presentation of Ellen van der Werff HERE and HERE

Supervised learning: Predicting passenger load in public transport

For many Public Transport (PT) users, overcrowding in PT vehicles has a major decreasing effect on the comfort experience. However, most online routing applications still not take comfort regarding to crowdedness into account, but provide recommendations based on shortest distance, shortest travel-time, or number of interchanges.
Being able to include certain information on crowdedness, requires knowledge about the current and future level of passenger load. Increasing amount and complexity of data describing public transport services allows us to better explore the detection methods and analysis of different phenomena of PT operations. Some countries or operators provide the possibility to use Smart Card (SC) data for occupancy prediction. However, SC data is not available in real time, which makes it hard to incorporate it into real time recommendation models. In this work, we show that it is possible to predict the passenger load via supervised learning, eliminating the need for fare collection data beyond the set needed for training.

Find the CASPT presentation by Léonie Heydenrijk-Ottens HERE

© 2011 TU Delft