Progress and Results
Summary of Deliverable 1.1: Big travel demand data analytics support tool
Deliverable 1.1 deals with the design and development of a tool to process, filter and analyze travel demand data. In several iterations of the requirements analysis process, the available data sets from the case studies were analyzed and the desired functionalities of the data analysis tool were defined together with all partners. Based on the requirements, the architecture of the tool was designed, and the tool was implemented accordingly.
The data storage backend consists of a modern graph database, which contains a transit layer and a passenger layer to store the cleaned and preprocessed input data. Queries against the database are implemented in the Apache Gremlin open source graph query language and can be called via a set of well-defined REST APIs. The APIs are defined using Swagger 2.0, which ensures easy documentation and integration into other applications.
Are you interested in the full text of this deliverable? Contact: Riccardo Scarinci.
Summary of Deliverable 2.2: Multi-layer Passenger Flow Network Model
Modelling transportation systems is a crucial aspect that allows the understanding of the internal passenger dynamics and the development of control strategies to improve the public transport service. This deliverable reports the modelling approaches for passenger flows and the related public transport operations used in TRANS-FORM.
We model three distinctive levels: hub, urban and regional. The hub-level model represents the dynamics, movement and activities of passengers in transportation hubs. We present a framework suitable to simulate, evaluate and generate pedestrian management strategies. BusMezzo, a multi-agent and multi-modal simulation, is used to represent the dynamics of public transport systems at urban level. The model includes transit services, travellers and management policies. The regional level is composed of the railway infrastructure and train traffic model. This includes origins and destinations of passengers, lines and stations, and train movement. The train traffic is simulated as a discrete-event model. The model is able to evaluate re-scheduling decisions on arrival and departures times for all trains.
Each level uses information from the other levels. The pedestrian travel times within a transportation hub, generated by the hub level, are used by the urban and regional levels to evaluate passenger dynamics more accurately. Similarly, the hub model uses information from the other two levels to generate precise passenger arrivals. The connection between the regional and urban levels takes place through the hub level. This means that hubs are the interfaces between levels. As a result, this modelling level is placed at the middle of the integration framework.
The models summarised in this deliverable are the foundation for the generation of multi-level control strategies for improving public transport services.
This deliverable is available here, D2.2-final
Summary of Deliverable 2.1: Passenger-focused Indicators of Service Variability
This document reports on the work performed in task 2.1 of the TRANS-FORM project, and is entitled “Passenger-focused Indicators of Service Variability”. This task was performed as part of WP2 “Measuring and Modelling Passenger Interchange Activities”.
The document discusses and proposes methodologies for quantifying and measuring service variability in the context of public transportation. To that end, the concepts of reliability and robustness are introduced. They describe service degradation in public transportation systems due to deviation from scheduled operations, considering recurrent disturbances and non-recurrent disruptions, respectively. A range of variability measures are presented that allow to analyse and evaluate passenger flows across entire public transportation networks. First, the state-of-practice is examined. Studied indicators include punctuality, quantifying the deviation of the actual arrival/departure time from the scheduled one, or regularity, describing the distribution of headways in high-frequency services. Advanced indicators such as residual capacity or connectivity reliability are also discussed. Such commonly applied variability indicators are typically limited to vehicle operations, i.e., they do not consider service variability from a passenger perspective. Particularly in the context of transfer flows, this represents a serious shortcoming. To address this limitation, a set of indicators is developed that assumes a traveller-oriented perspective. These range from extensions of existing measures, such as passenger-weighted vehicle punctuality, to novel service- oriented vulnerability indicators that are to be practically assessed and established for the first time within the TRANS-FORM project. Examples include likelihood of missed connections, route-travel time variation, passenger-specific excess travel times and their variability, or perception-based journey times. The proposed indicators are first discussed and developed separately at the level of a transportation hub, at the urban and the regional level, before they are combined in an integrated framework.
The aforementioned measures are tested and applied to three different case studies that are considered throughout the TRANS-FORM project, allowing to understand their nature at the various aggregation levels and across national borders. The wider aim of the development of service variability measures pertains to the development of a multi-layer passenger flow model that can represent flow dynamics across a public transportation network, and which allows to evaluate and steer transit operations from the perspective of individual passengers. Specifically, in the subsequent work packages of TRANS-FORM, the proposed indicators will be used within a traffic management tool to evaluate system performance under current or alternative scenarios, for instance in terms of disruptions, and to develop management strategies that minimize them.
This deliverable is available here, D2.1-final
Summary of Deliverable 3.1: A toolbox of real-time strategies for smart transfers
Organizing, financing and operating public transport service networks can be quite a challenge. The liberalization of the public transport sector within EU has introduced some additional challenges since the public transport systems nowadays more often consist of services operated by multiple organizations. When passengers move between public transport service networks that are operated by different organizations, the need for effective coordination become evident. The importance of effective coordination and the associated challenge to achieve this – independent of organizational structure – is the point of origin for the TRANS-FORM project. The project focuses on the development of an integrated passenger-focused management approach that takes advantage of multiple data sources and state-of-the-art scheduling support.
This document reports on the work performed in task 3.1 entitled “Real-time traffic management optimization” and task 3.2 entitled “Smart real-time strategies” of the TRANS-FORM project. These tasks were performed as part of work package 3 “Methods for Planning and Operating Robust Services”.
The work in the mentioned tasks focus on how to model an integrated passenger-focused management approach including developing strategies that enable improved coordination and smooth passenger transfers. This document contains a specification of the configuration of each modelled level (hub, urban and regional) and the proposed types of management strategies as well as the required information flow. A specification of the proposed integration of those three levels and their interaction is also presented.
Are you interested in the full text of this deliverable? Contact: firstname.lastname@example.org
Summary of Deliverable 5.1: Case Study Set-up Descriptions
The three main chapters of the document covers the three case studies covered by the TRANS-FORM project. First, the hub level case is presented. For this case, the application are pedestrian movement analysis within the train station in the city of Lausanne, Switzerland. Second, the urban level case is presented. In the urban level case, The Hague is used as the case study area. The focus is on modelling the public transport flows operated by HTM at the urban-metropolitan level in this area. Third, the regional level case is presented. This case use the Blekinge region, Sweden as the case area. The focus is on regional trains and busses, organized by the regional public transport authority, Blekingetrafiken.
This deliverable is available here, D5.1-final
Summary of Deliverable 3.2: Algorithms for short-term prediction of passenger flows
Public transport networks are increasingly complex and multimodal and thus challenging to manage, while it is increasingly important that the networks are robust. One of the goals of this project, TRANS-FORM, is to develop effective strategies and methods to support decision makers in transforming public transportation into a seamless travel experience perceived well by the travellers. Furthermore, the project focuses on the development of an integrated passenger-oriented management approach that takes advantage of multiple data sources and state-of-the-art scheduling support. The “passenger-oriented” means that the planning and operational control of the public transport service network do not only concern operator- and network-related aspects, but also considers the reliability of passenger transfers and quality of service. In order to introduce and apply effective passenger-oriented control and management strategies, predictions of the passenger flow dynamics – including passenger route choice estimations due to real-time re- scheduling of services – could be useful.
This document reports the work performed in task 3.3 entitled “Short-term forecasts of passenger flows” as a part of work package 3 “Methods for planning and operating robust services” of the TRANS-FORM project. The work in the this task focuses on how to predict short-term demand, as an effect of changes in the transport service network due to e.g. disturbances and passenger preferences. This document presents the algorithms for passenger flow predictions to be incorporated in the real-time management strategies for the three different levels (hub, urban and regional) that are outlined in the previous project deliverable “D3.1: A toolbox of real-time strategies for smart transfers”.
Are you interested in the full text of this deliverable? Contact: Yuki Oyama.