Ongoing Projects
STREET: Simulating Transportation for Realistic Engineering and Education. National Science Foundation
Cost: $500,000. Time: Four years.
Currently, most undergraduate transportation engineering courses are offered in a traditional combination of “chalk and talk” lectures and “paper and pencil” problem-solving exercises, failing to fully motivate students and preventing them from effectively assimilating and applying knowledge in their future work. This approach to education does not expose undergraduate students to the myriad of challenging issues and solution methods that would encourage them to pursue careers in transportation engineering. A new paradigm for transportation engineering education is needed to better engage students and deliver knowledge. Although simulation has been proven to be powerful tool to encourage active learning, the effectiveness of simulation has not been fully examined in transportation education practice, and its potential advantages over traditional ways of learning have not been widely acknowledged. The focus of this proposal is to develop and refine a suite of web-based simulation modules that can be easily incorporated in the undergraduate transportation courses. These simulation modules will be also evaluated in a diverse setting and improved after examining evaluation results. Eventually, the simulation-based teaching materials will become an active textbook, which offers an interactive learning environment to undergraduate students. The active textbook with simulation is expected to improve student understanding of critical concepts in Transportation Engineering and student motivation toward Transportation Engineering and improve student retention in the field. Specifically, targeting the required undergraduate course Introduction to Transportation Engineering, five simulation programs are in the development plan, namely, ROAD: Roadway Online Application for Design, OASIS: Online Application of Signalized Intersection Simulation, SOFT: Simulation of Freeway Traffic, ADAM: Agent-based Demand and Assignment Model, and SONG: Simulator of Network Growth. Prototypes of these simulation programs have been developed (including ROAD, ADAM and SONG) or currently being developed (including OASIS and SOFT) through the PIs’ accumulated works over the last several years. Initial evaluation and testing including the prototypes of ROAD, ADAM and SONG have been conducted in the course offerings at the University of Minnesota - Twin Cities (UMN). The preliminary evaluation results are very positive and encourage us to pursue further. This project will further develop and refine the five simulation modules and integrate them into the undergraduate courses at UMN. Building upon the local implementation success, the simulation modules will be evaluated and tested in the course offerings of over sixteen other transportation programs from different universities across the country. Feedback from the implementation will be provided to the project team for continuing improvement. In addition, to reach an even broader audience with computer simulation, a mobile booth with public-oriented versions of some of the simulation modules will be displayed at the annual Minnesota State Fair and the Minnesota Transportation Museum for public demonstration and education on current transportation problems.
Small Grant for Exploratory Research: Responding to the Unexpected: Understanding Travelers' Behavioral Choices in the Wake of the Mississippi River Bridge Collapse. National Science Foundation
Cost: $17,000. Time: One year.
This research aims to (1) Collect both aggregated traffic data and disaggregated panel survey data that will help to understand the traveler’s behavioral choice and traffic pattern evolution after the event; (2)Perform preliminary analysis on collected traffic data; and (3) Identify short-term and long-term research needs concerning traveler’s behavioral response and traffic management strategies after network disruption.
Economic Impact of Upgrading Roads. Minnesota Department of Transportation
Cost: $72,636. Time: One year.
Improvements to transportation networks, especially those in growing areas, tend to have impacts on local land markets. In principle, an improvement to a link in the network will confer economic benefits to adjacent and nearby properties. Depending on the type of improvement (construction of a new link, capacity addition to an existing link, or upgrading an existing link), the benefit could represent a reduction in the time cost of travel or other variable costs (fuel consumption or mileage-related vehicle depreciation). Theory would suggest that these benefits are capitalized into local property values, yielding a localized external benefit. This research will explore the nature and magnitude of such benefits. Using a sample of property sales data for Minnesota counties, both rural and urban, we will explore the impacts of upgrading roads on nearby property values. We will also attempt to track and model the cost of road upgrades considered in the sample in order to provide rough comparisons of these costs with the estimated benefits flowing from the various projects. It is also important to note that such upgrade can have a negative effect due to the external factors (e.g. noise, pollution). These factors will also be taken in the estimation of the effects of network upgrades on nearby land value.
Post-Construction Evaluation of Forecast Accuracy. Minnesota Department of Transportation
Cost: $51,832. Time: One year.
Evaluations of proposed projects and their subsequent performance depend critically on forecasts of demand made well in advance of actual implementation. Depending on the forecasting method used, there may be considerable uncertainty in projections of future benefits or other performance criteria. The current study seeks to evaluate the accuracy of demand forecasts made for a sample of recently-completed projects and estimate the impact of forecast error on post-construction measures of project performance. Data for the study will be drawn from completed environmental impact assessments, which will provide pre-construction demand forecasts, combined with recent traffic volume data from Mn/DOT. Results can be compared for urban and rural projects, as well as for different functional classes. The research will conclude with some suggestions for managing and possibly reducing uncertainty in project demand forecasts.
CAREER Award: The Evolution of Transportation Networks: Empirical Research and Agent-Based Models. National Science Foundation
Cost: $400,000. Time: Five years.
This research endeavors to understand the evolutionary growth process of transportation networks at a theoretical and empirical level, recognizing the inter-dependence of supply and demand, and to develop agent-based models to replicate that process.
Access to Destinations: Monitoring land use activity changes in the Twin Cities Metropolitan region. Minnesota Department of Transportation and Hennepin County, Minnesota.
Cost $88,000. Time Two years.
Any accessibility measure has two main components. First is land use activity or attraction, while the second is transportation system including network impedance. This research concentrates on changes in land use activity over time in the Twin Cities metropolitan region as a part of the Access to Destinations research program being developed at the Center for Transportation Studies, University of Minnesota. In a typical land use and transportation cycle, any change in land use is expected to lead to changes in the transportation characteristics. Meanwhile the change in transportation infrastructure and/or characteristics is expected to lead to changes in land use and so on. Accordingly it is essential for any accessibility project to have a full understanding of the consequences of changes. This research propose to develop a land use change prediction model that can be integrated with network growth models to help in developing measure of accessibility over time. This model will assist in understanding the consequences of changes in land use activity in the Twin Cities metro region while accounting for changes and effects of the transportation system.
Job-Housing Mismatch: An Affinity Model of Worker-Job Matching. Center for Urban and Regional Affairs University of Minnesota.
Cost ~$44,000. Time One year.
Traditionally, travel surveys based on 1% of the population are used to understand travel behavior. Recently, much more complete data has become available to researchers and policy makers, allowing them to better understand job- housing mismatch in a region. The Longitudinal Employer–Household Dynamics dataset, available from the Minnesota Department of Labor, includes residence and employment location information for each employed individual in the Twin Cities metropolitan area (excluding self-employed and selected sales personnel). Using these data, as well as data from other sources, this project will analyze the relationship between people’s choices of residence relative to their employment locations in the Twin Cities region.
This project will extend travel behavior research to help understand the characteristics of people’s choices of residence relative to their employment locations. The central research hypothesis is that although travel time and income are important factors in where people live and work, other factors may help shape the commuting patterns observed in metropolitan areas. Certain residential neighborhoods produce more workers for a given employment district and in a given industrial classification than can be explained by travel time and income alone. By better identifying the causal factors in travel location, travel demand modelers, transportation planners, and engineers will be better able to address job-housing mismatches and imbalances between demand skills and worker skill sets.
The role of Social Networks and Information and Communications Technology on Destination Choice. Humphrey Institute of Public Affairs State and Local Policy Project, Techplan Program.
This research proposes to investigate the impact of traditional social networks and information and communications technology (ICT) on travelers’ destination choice. Earlier models of travel demand were aggregate in nature and found their roots in statistical mechanics rather than the behavioral basis of those decisions. Over the years advances in disaggregate modeling have led to improvements in traditional models giving them theoretical framework that related human behavior. As time has progressed, so have the models of travel demand going from the traditional four step process to disaggregate models of individual choice as widely seen in route choice models. These models by and large assume independent actors that make decisions based on their own personal taste. The extent to which social networks and information and communication technologies affect where our destinations are located is an area that is gaining more focus. This project proposes to undertake a study of the extent to which both traditional social networks and communication technologies (and social networks mediated by those technologies) affect these choices. In two phases this project hopes to develop models that incorporate important elements of social networks and ICT for different trip purposes.
One dimension of travel is that part of it is derived to engage in activities where other people, who can be friends or acquaintances, are met. Even the work trip can be thought of as an engagement where multiple people get together in a particular time slot to engage in their respective responsibilities. It has already been established that a majority of work finding occurs through weak social ties rather than through any formal search (Granovetter 1974). Recent work by the authors of this proposal on a randomly selected area of the Twin Cities observes that there is a statistically significant clustering between where individuals live and where they work. One possible explanation for this clustering is that some of the networks that lead to finding jobs may be transmitted at the neighborhood level leading to the observed clustering.
The Internet also provides a wide range of information on job availability. It is expected that a number of people find their jobs using this medium, but also that there maybe differences across sectors of the economy on its prevalence o as a tool for finding jobs. Knowing the demographic characteristics of the people who use this medium to find work, and which sectors are more likely than others to find workers using this medium allows more specific analysis than knowing only how many workers are attracted to a particular zone.
Outside of the work destined trip, it is also important to ascertain the extent to which ICT affects destination choice. Just as the Internet can be used to find jobs or organize meetings for multiple people, it is also possible that expanded access to the web has replaced the demand for face-to-face meetings and/or the requirements of travel for some activities. For example, people who want to buy a book can get more value by replacing the shopping trip to an hour on the web where they can find more information on a book from others that have read it before and more recommendation. In other cases the web can also be used to locate particular destinations for leisure, meetings or shopping. It both expands the choices for people beyond what they know of locally and can potentially replace some activities.
To answer these questions on the role of traditional social networks and ICT on destination choice, this research proposes to undertake a detailed study of employees of a particular firm where individuals are asked about the mechanisms through which they located their jobs. For example we will ask what means (e.g. recruiters, informal contacts, the web etc.) they used for finding their jobs; if contacts are used, we will ask where they are located and how information about the work reached them. By understanding the information path for different types of jobs and the systematic similarities between them, it is expected that the models of work destination can be made more specific than those done at the aggregate level.
For non-work trips, respondents will be asked how their social networks and the web affect their decisions for destinations. Some of the questions of interest are to what extent they organize meetings (both physical and virtual) over the web (limiting their network to those who have similar access), do they replace trips because of the web, do they find new destinations using the web, how in general does it affect their choices of where to shop or undertake leisure activities? Again it is expected that these considerations would lead to better and improved models that take into consideration that the people to be involved at those activities affect the destination choice in addition to the activities themselves.