Ideas for thesis topics

Commonly, we expect student to develop their own topics for a thesis, and we are available to brainstorm on potential topics that fit the students interests and skill sets.  We also post topics to this page that we feel are interesting for a thesis. 

Time for Activities

One of the most important driver of travel behavior is the amount of time we spent on various activities at different places. Typically, we use the household travel survey MiD to explore activities. We know, however, that many activities are underreported (such as the visit at the bakery might not be reportet because it feels to be too short to be relevant or because it was simply overlooked). Therefore, it is important to compare the amount of time spent on various activities and the resulting travel across several data sources.

A valuable data source that is currently underused in travel behavior analyses are time use studies. The German time use study provides detailed data how people spend their day. This thesis will analyse the amount of time spent on various activities and travel from three data sources: The German time use study, the MiD household travel survey and the MOP mobility panel dataset. The data will be provided to the student. Finding (in)consistencies will be an important step to better understand and explain travel behavior. 

Automatized link volume estimation using open data

With the advances of the digital world and the willingness to create and share more and more data sets collaboratively, new opportunities for data analysis and use have also emerged for research. For some applications, such as assessing the environmental impact of traffic, traffic volumes are needed. In this thesis, we will develop a tool chain that estimates daily traffic volumes per road using widely available free data, such as OpenStreetMap or facebook data for goods. The goal is to process traffic volumes for noise analysis, which is why characteristics such as trip purpose, origin-destination relation, etc. are irrelevant. As a classical input-output problem, a machine learning approach is preferred, which is transferable to the whole of Germany if possible. One focus of the work will be to evaluate the quality of the model with respect to large-scale application.

Impact of access restrictions for diesel vehicles in Munich starting in 2023

Starting in 2023, Munich will restrict access for older diesel vehicles in an attempt to reduce emissions, particularly particulate-matter emissions. While this move appears to be warranted from an environmental point of view as well as from a public health perspective, the social equity implications could be significant. In this study, it shall be analyzed with an agent-based simulation model who is affected by this ban from older diesel vehicles. The model needs to be expanded to distinguish vehicle types. A statistical analysis shall define who is most likely to own vehicles that will be banned from entering Munich. The equity analysis will explore to which degree which segement of the population is affected, and whether these people have transport alternatives other than buying a newer car. The topic requires some interest in working with models, including some programming in Java and statistical analysis of car ownership data. 

Impact of weather on travel behavior

Travel is influenced by weather. Rain, fog, snow, wind and particularly high or low temperatures influence whether people leave their home, what time of day they travel, how far they travel and which mode they choose. In this research, weather data shall be attached to a household travel survey to identify which weather events affect to which degree different aspects of travel behavior. Statistical analyses may include regression analysis, discrete choice analyses and/or machine learning approaches.

Ride-hailing pricing and public transportation

Since about a decade ago, transport network companies (TNC) have been prevailing all over the world. Starting from Uber in San Francisco, Lyft, Grab, Didi, and Cabify have launched in different regions. Enormous amounts of trips are made, and it is estimated to be up to 10-20% of all vehicle miles traveled in the city centers. This massive growth once brought the expectation that it could be a supplementary mode of public transportation, yet the actual measures to encourage it are left to be observed. Integration of the two modes is needed so that they are more complementary instead of competitive, and pricing can be one simple approach to make it happen. (Skills: Python, R, or coding skills in other languages)

Contact: Zhihua Jin

Equity and profitability in public transportation fare structure

From the old complicated fare structure, MVV introduced a 6-zoned fare scheme for public transportation in 2017. Zone-based fare has been on trend since the 1980s; after London shifted from distance-based fare to zone-based fare, Zurich, Berlin, and many other European cities followed. More than 20 years have passed since most of these cities implemented this fare structure. With the introduction of Deutschland tickets and some MaaS platforms in sight, is it still a comprehensive fare structure in this subscription era? How does spatial equity work now? A contemporary analysis is needed to study the difference between subscription fares and zone fares for passengers as well as operators. (Skills: Python, R or coding skills in other languages)

Contact: Zhihua Jin