Projects, which are under progress or already confirmed (funded, assigned, etc.)
MATSim model for Santiago de Chile
VSP - TU Berlin
Universidad de Chile
Creating a MATSim scenario for Santiago de Chile from open data
The first version of the scenario is now available online. It includes a car network based on OSM data, a population based on a detailed travel diary survey (~0.65% of the total population), and public transport supply data from GTFS which runs on a separate network.
|MINTE (http://www.tu-berlin.de/index.php?id=146151)||VSP - TU Berlin||Optimal combined pricing for emission and congestion|
We have tested simultaneous pricing for emission and congestion external costs on (a) Sioux Falls and (b) Munich, scenarios and compared the results with individual pricing schemes.
Optimization and network wide analysis of traffic signal control
VSP - TU Berlin,
Analyzes the interaction between traffic signal control and users’
behavior from two perspectives: optimization of traffic assignment and signal control in a mathematical model (by BTU Cottbus) and evaluation of the impact of signal control optimization on travelers’ behavior using MATSim (by VSP).
Currently analyzing simulation parameter and signal influences in small scenarios like Braess's paradox and the inner city of Cottbus, especially regarding route choice and travel time.
|Optimal Pricing Strategies|
Investigating the use of simulated dynamic pricing to optimize transport systems; Simulation-based computation and internalization of external effects
|Finalizing a study where tolls are simultaneously set based on congestion and noise cost. Investigating and comparing different congestion pricing approaches.||Ihab Kaddoura|
eCab: Simulation-based system for the sustainable management of a fleet of electric taxicabs
VSP - TU Berlin
|The goal is to (a) investigate possible ways of introducing electric taxicabs in Berlin, (b) devise efficient strategies for dispatching and charging taxis, and (c) estimate the outcomes in terms of sustainability.||Online electric taxi dispatching and charging as the assignment problem" (by Maciejewski and Bischoff) presented at the 28th European Conference on Operational Research, EURO2016.|
A new EV model (for the Nissan Leaf) for both driving and AUX-related energy consumption has been added to MATSim (currently, in the michalm playground).
|February 2013 - January 2016|
|IRP project||BUTO - Vrije Universiteit Brussel||Intermodal Freight transport model to calculate external costs.|
|Initial demand from mobile phone trajectories||VSP - TU Berlin||Creating Initial demand from mobile phone trajectories|
Take a calibrated MATSim run; Put a mobile phone network coverage on top; Give agents a calling behavior; Collect an artificial mobile phone dataset.
Set up a new scenario with only the artificial mobile phone dataset and link counts to use: Create an expanded population, have each agent draw from possible realizations of its individual mobile phone trajectory; calibrate with cadyts. Compare to ground truth.
|Catch-a-car||IVT - ETH Zürich||Scientific supervision of a free-floating carsharing project in Basel. New mode choice models, with FF carsharing as alternative, will be estimated and implemented in MATSim.||The scenario for the city of Basel is being generated||October 2014 - October 2016|
|Valuing Travel Time||IVT - ETH Zürich||Aiming at expanding existing models for the number, type and sequence of out-of-home activities, elaborating time use, daily scheduling, stated choice results and value of time components for the region of Vienna and evaluating the resulting policy implications using an updated version of MATSim.||2014-2016||Basil Schmid|
|Post Car World||IVT - ETH Zürich||Analysis of the mobility patterns in a non-car oriented world. Induced/suppressed demand estimations using surveys and MATSim scenarios.|| |
Third way of survey completed. Data clean-up and modeling are under way, followed with the creation of MATSim PostCarWorld scenarios
|A generalized approach to population synthesis||IVT - ETH Zürich||Create a robust and fast population synthesis tool to improve the overall quality of MATSim results.||The population is available, tests are being run.||20??-2016|
|SCCER Mobility||IVT - ETH Zürich||In the first year we worked on Electric Vehicle Infrastructure pricing and are currently working on modelling trade-offs between Electric Vehicles and Conventional Vehicles using MATSim.|
(could be extended 3 additional years).
VSP - TU Berlin
IVT - ETH Zürich
|The Book is available!||2014-2016|
|Sioux Falls||FCL Singapore||Extension of the Sioux Falls scenario with a finer OSM network including public transportation and improved accessibility. Addition of through and freight traffic in order to create congested traffic condition during peak hours.||Sebastian Hörl|
|User preference heterogeneity (Economic Valuation)||FCL Singapore||In depth analysis of effects of value of time heterogeneity and different types of schedule delay heterogeneity among users. Comparison to the transport economics theory using a simple multi-modal corridor scenario.|
Review and redesign of the code from my PhD Thesis "Urban Mobility Pricing with Heterogeneous Users" with the goal of creating a MATSim contribution. The new Contrib should enable convenient inclusion of value of time and schedule delay heterogeneity into the simulation based on individual scoring and routing parameters for each agent.
|Secondary activity scheduling||FCL Singapore||With established mandatory activities in people’s daily plans (rest, work, study), time planning, scheduling is primarily based on organizing flexible activities and trips during the given remaining time windows. This work develops a method to extract and model flexible activity patterns from real data, and to use them to make a massive number of activity chain predictions.||Sergio Ordonez|
|Distributed simulation||FCL Singapore||A master-slave configuration for MATSim, where repeated replanning and plan evaluation of very large scenarios takes place on cheap slave nodes, passing plans to a master node executing agents' selected plans in QSim, and receiving TravelTimes from the master to update the meta-model (PSim) used for plan evaluation and scoring. Works on ethernet and StarCluster (Amazon EC2)||Pieter Fourie, Sergio Ordonez|
|Transit smart-card-driven MATSim||FCL Singapore||Demand derived from transit smart card data drives a simplified simulation of transit operations, where stop-to-stop speeds are sampled from the (inferred) speeds experienced by buses in the actual network, eliminating the need to simulate car traffic.||Artem Chakirov, Alex Erath, Pieter Fourie, Sergio Ordonez|
|Activity-based Modeling Using Smart Card Data||Department of Transportation Engineering, University of Seoul|
In this work, Transit demand is generated by exploiting the smart card revealed daily transaction records. In the first step, trip segments are converted into trip chains followed by the INFERENCE OF ACTIVITY TYPES. A set of rules are defined by using socio-demographic profile (Household Travel Survey Data, 2010) to assign activity purpose. Daily itineraries of agents' are then converted into MATSim demand and calibrated with HHTSD.
|Waste collection in Cape Town||ISE - University of Pretoria||Trying to model waste collection vehicles in MATSim based on actual GPS data||Using map matching techniques in order to create MATSim routes for a subpopulation of waste collection vehicles|
|SMART-PT – Smart Adaptive Public Transport|
VSP - TU Berlin
DTS - Poznan University of Technology
|European long-term project. Systematically analyze the discrepancy between end-user travel demand and existing urban transit supply. Uses minibus contrib.||Working on large-scale autonomous taxi (AT) services and assessing their impact on public transport, traffic and charging infrastructure (the latter for electric AT scenarios)||2014 - 2016|
|Machbarkeitsstudie "Teilszenarien und Standardbusse"||VSP - TU Berlin||Application of the minibus contrib to 10% scenarios, scenarios with full-size buses, and huge scenario areas.|
|MAXess - Measuring accessibility in policy evaluation|
ISE - University of Pretoria
VSP - TU Berlin
IVT - ETH Zürich
|Calculate accessibilities (taking into account both land-use and transport system characteristics) at a disaggregate level based on freely available data and make results available through high resolution, open-source, adaptive maps to allow the assessment of planning schemes.||Our geoserver is now set up and hosts initial results; see http://geo.vsp.tu-berlin.de|
|Sharing is saving - NFP||IVT - ETH Zürich||Solutions to limit the negative impacts of transportation, especially the reduction of energy consumption. It focuses on collaborative mobility systems. Scenarios assuming their further diffusion are set up and evaluated with MATSim||Scenarios are being created based on a literature review, first assessments with MATSim are being run||2014-2017|
|Autonomous Vehicles - SNF||IVT - ETH Zürich||Quantitatively study with MATSim what the arrival of autonomous vehicles will mean for the transport system and demand. Provide libraries to simulate autonomous cars in MATSim.||Creation of possible future AVs scenarios based on qualitative evaluations and expert knowledge||2015 - 2018|
|Integrated population synthesis and demand generation of large-scale scenarios||FCL Singapore||A project for the Urban Redevelopment Authority of Singapore, to rapidly generate future population scenarios, along with resulting MATSim activity chains and travel demand.||Alex Erath, Pieter Fourie|
|MATSim model for Stockholm||KTH Stockholm||Work funded by two different projects. One is SMART-PT, supported by the EC and focusing on a model of the public transport system, and the other one is IHOP3, funded by the Swedish Road Agency and aiming to integrate the different existing Swedish demand and supply models in a flexible manner, using MATSim as a middle layer||KTH is currently building a MATSim model for Stockholm,||Gunnar Floetteroed|
|Traffic behaviour in all-new city districts within metropolitan areas||SI (Städtebau-Institut) - Universität Stuttgart||Alteration of the urban mobility system when newly-built residential or industrial areas are connected to the transport system using the example of Stuttgart. About the correlation of mobility and urban structure on different spatial email@example.com|
|Transit Optimization||VSP - TU Berlin||An agent-based evolutionary approach for the user-oriented optimization of complex public transit systems http://www.vsp.tu-berlin.de/projects/laufende_projekte/transit_optimization/.|
|Estimating a Germany-wide OD-Matrix for private car travel - Update 2014||Deutsche Bahn||The aim of this project was to estimate a Germany-wide OD-matrix on a zonal structure that is comparable to the German zip codes. The OD-matrix was generated as a aggregation of a microscopic matsim scenario based on the German travel survey (MiD 2008) and calibrated against 2500 count stations.||We extract a fine-grained OD-matrix from a Germany wide synthetic population describing average daily travel patterns. The 2014 update includes more and update data sources as well as methodological improvements in the imputation and calibration process.||2015|
|The influence of parking availability on Mode choice behavior and car ownership||VSP - TU Berlin|
In a simulation based study of two districts in Berlin, mode choice behavior is evaluated depending on the available parking space in the area. For this, a within-day-replanning based parking simulation is integrated. Effects of dedicated free floating carsharing parking infrastructure will also be assessed.
|Parking space limitations are currently integrated into our Berlin scenario.||Nov 2016||Joschka Bischoff|
|VSP - TU Berlin||Using an extension for MATSim's DVRP algorithm, we are implementing a system of shared taxibusses for collective transport that could e.g. be organised by employers of large industrial sites or in rural areas. A scenario has been set up, the extension for DVRP has been implemented, though the optimizer is still in a very early stage.|
The optimizer is running for a specific case study and achieves results that are similar or better compared with schedule based pt. Results are pre-liminary and non-public, but publication is expected at a later stage for a different scenario.
Expected Project end: March 2016
|Economically consistent simulation of travel behavior with MATSim (IHOP3)||KTH Stockholm||The question addressed in the present project is if and how it is possible to perform consistent CBA with this model system.||We have been developing a strategic transport model system for Stockholm. An uncalibrated but functional prototype is now available.||Gunnar Floetteroed|
|Model Germany||senozon||Together with microm (http://www.microm-online.de), Senozon builds a multimodal MATSim model of all of Germany.||After first building a test model for the region of Nuremberg in 2015, a model covering all of Germany was built this spring. 10% runs are finished, 25% simulations are currently run. The model exists in two variants, one using a network based on OpenStreetMap, and one based on a commercial navigation network.||2015-16|
|Model Austria||senozon||Senozon is building a multimodal MATSim model covering all of Austria.|
Data availability has been assessed in January and February 2016. Work to build the model started in late February and is currently ongoing. We plan to have the model ready in late-summer 2016.
Project that one plans to do or which are hoped to be funded.
|Mitigating negative transport externalities in industrialised and industrialising countries (add to existing)||Queue model is extended to depict the behaviour in mixed traffic conditions||March 2017|
|New Data for Social Leisure Travel Forecasting||A data collection effort to improve models of joint leisure activity involvement, and build those models into MATSim to test their ability to improve the forecasts of leisure trips.||2016-19||Thibaut Dubernet|
Project ideas which are dropped this month (because of rejected funding, lack of interest, etc.)
Information about current work on contributions which are not directly related to a formal project
|I have been developing a new method and the corresponding code to rather quickly approximate the solution of rather general optimization problems where evaluating the objective function requires to run a simulator like MATSim. The code is functional and currently located together with an example in my playground. Please get in touch with me if this is of interest; I think it would make sense to accompany the move from playground to contrib with a concrete use case. I can also make available a working paper describing the method.||Gunnar Floetteroed|
Events/Notes/Information that are Worth Mentioning