Education

CIE4811-18 Planning and Operations of Public Transport Systems

This is an introductory course to the planning and operations of public transport systems. Students will learn how public transport systems are planned, starting from the long-term strategic planning, going through tactical planning and finally discussing its real-time operations. Planning dilemmas and solution approaches to planning topics prevalent in passenger transport systems will be discussed and applied. Lectures are given by a team of teachers.

CIE5825 Advanced Public Transport Operations and Modelling

This is a follow up course of CIE4811-18 (Design and control of public transport systems). It addressess more details and insights into public transport operations and modelling. The following topics will be covered: - Determinants of public transport mode and route choice - Principles, approaches and tools for transit assignment - Design principles and dilemmas for demand responsive public transport - Control measures (eg headway control, rescheduling, holding, etc) The main goal of the course is to evaluate and debate public transport operations and modelling, deepening the knowledge of students on efficient operations and control and modelling (both strategic and tactical)

Training

Research Opportunities

MSc Thesis: City-wide benefits of shared taxi rides

Modern taxi services, also known as ride-hailing services, like Uber and Lyft offer in addition to individual rides also shared rides, where travellers pay less while allowing for longer travel and/or waiting time. Pooling passengers into shared-rides is believed to be a game-changer in modern mobility as a service (MaaS). There are high expectations on reducing congestion and externalities of car traffic through providing door-todoor alternative service at a low price. At the same time, the potential for pooling rides and the additional empty-vehicle trips it may induce, remain largely unknown and call for the development of new methods and experiments. Recently proposed methods allow to match trips into attractive shared rides, applicable also for large sets of urban trips. Single OD trip requests are effectively matched into shared rides, attractive due to price discount compensating for the discomfort of sharing. Resulting shared-rides may be further analysed and quantified to understand the actual scale and potent of trip sharing benefits.

MSc Thesis: Simulating MaaS drivers’  behaviour with Agent-Based-Model

In a two-sided mobility markets where service is provided by self-employed drivers paid by commercial
platforms (like Uber, or ViaVan drivers), the supply of transport services is an outcome of drivers’ choices.
MaaS service providers, unlike conventional tendered public transport, are not designed or controlled, but are
rather the collection of choices made by individual agents.
Each driver is free to accept or decline trip requests and operate in areas that s/he finds attractive. They
have individual strategies to maximize their profit (e.g. reposition to attractive areas such as when surge
pricing is active) and select the service they provide (individual or shared rides). Moreover, drivers may
independently decide on their shifts and working days, including possibly changing job and stop providing
services at all.
This makes the picture of MaaS supply side challenging to analyze, predict, describe and optimize. It is still
unclear how to best represent drivers’ behaviour and how it affects the system performance and reliability.
In this project you will address this research gap with simulation experiments.

MSc Thesis: Spatial and social patterns of passenger demand in Amsterdam

The master graduation project will involve the following activities, all in relation to the case of the public transport system in the Amsterdam region

  • Computing public transport accessibility indicators
  • Analysing origin-destination demand patterns
  • Segmenting the demand patterns in relation to key socio-demographic variables
  • Investigate the relation between service accessibility and socio-demographic characteristics
  • Consider implications for fare policy strategies

MSc Thesis: Unravelling Passenger Demand Patterns using Smart Card Data

Smart card data from the Stockholm region is available for this project. Passengers tap-in on buses or at station gates in the case of metro and commuter train. No tap-outs are available but they have been inferred in a previous project, making is possible to estimate OD matrices and travel patterns. The master graduation project will involve the following one or more of the following activities, all in relation to the case of the smart card data in the Stockholm region: Analysing origin-destination demand patterns including spatial clustering; Segmenting passenger demand based on travel patterns; Test the tap-out inference technique by calibrating it against counts; Investigate the impacts of a major network change – new commuter train stations and frequencies - on passenger travel patterns.

MSc Thesis: Impact of reliability and crowding on public transport users’ choices

The purpose of this project is to find the impact that reliability has on travellers’ public transport alternative choice. To do so, a choice experiment of stated preferences will be carried out. In this experiment, each alternative will have different values of speed, frequency, headway regularity and average occupation. This experiment has already been designed and is currently conducted in Santiago de Chile. The attribute values will have to be determined for the Dutch context (e.g. Amsterdam metro or Dutch Railways) and a series of choice model estimations will be performed to establish the trade off between variables and the value of reliability. Results will be compared to findings from Santiago de Chile where crowding is prevalent.

MSc Thesis: Assessment of On-Demand Transport

The masters student will work on the agent-based simulation platform(MATSim) and assess the impact of various operational aspects of demand responsive services (door-to-door/stop-to-stop, private/shared, paratransit, and car-sharing) on the mobility of users. The impact of these services will be analysed in terms of user perspective (generalised travel cost), operators perspective (operating cost, revenue), and system perspective. The model will be tested on the fully developed agent based simulation scenario of Amsterdam.

MSc Thesis: Analysis and mitigation of bunching effects on metro services

The strategic and tactical planning of public transport networks require advanced tools to model, analyse, quantify, optimize and evaluate current and alternative service designs. Challenges include the interactions between its infrastructure layer, service layer and passenger flows, as well as the importance of behavioural and dynamic phenomena. Projects in the domain of network analysis, service dynamics and service optimization can be performed at the rail group of Royal HaskoningDHV, an international engineering consultancy firm.

Smart Public Transport Lab

 TU Delft logo White

Contact:
 Phone: +31 (0)6 15908644
 E-mail: smartptlab@tudelft.nl

Address:
 P.O. Box 5048
 2600 GA Delft

Visiting address:
 Building 23
 
Stevinweg 1
 2628 CN Delft