After spending time in the US, I am amazed how much one has to drive and this got me thinking about using agent-based models for traffic simulations (which is a large body of literature accompanying it). It also relates to my interests in urban systems and the fact that as cities have grown, transportation technologies have evolved (from walking, to trains etc), and now the automobile has become the dominant mode of transport for moving within and between cities. Trips range from journeys to work to shopping trips. The wide spread adoption and use of automobile is also one of the contributors to sprawl (in its many shapes and forms) as the car is not restrained by frequent stops or set routes, for instance such as trains are. Thus, if one can understand the relationships between land use and transportation one can investigate issues relating to urban sustainability. This is where agent-based models come in, in the sense they allow one to focus on the behavior of people. For example, how people decide to go to work.
ABM also allows us to explore simple thought experiments and how more aggregate results emerge from individual interactions such as: what is more effective, a four way stop or a traffic lights at a road intersection? The simple agent-based model presented below utilizes MASON and was created by Omar Guerrero of the CSS department at GMU. The rules of the model are simple, in the sense that at a four way stop, the vehicle that is first to arrive, it is first to move, unless two vehicles arrive at the intersection at the same time and then the vehicle has to give way to the car on the right. While at traffic lights vehicles must stop at red lights. The movie below shows part of the graphical user interface for a particular model run of both a four way stop and traffic light.
Even though this is a simple model, one can explore a number of issues such as how these different intersection configurations impact on the flow of traffic under different volumes of traffic. For example, at low traffic volumes in general, the stop sign is more effective (i.e. allows more cars to cross) than the traffic light. However, at greater traffic volumes the traffic lights out performs the four way stop (in the sense there are more cars in queues) but also with high traffic volumes, one can see oscillations in traffic waiting at the lights while the four way stops create long queues of traffic as shown in the figure below:
Moving away from micro patterns of traffic flows one can use ABM to explore daily commuting. For example, traffic models such as TRANSIMS or MATSim allow for the study of entire metropolitan regions and how traffic jams etc. form. To give a simple example of such a movement, the model presented below illustrates how many individuals can cause traffic jams. The model (using GeoMason) is based on commuters working within the Tyson’s Corner area of Virginia which boarders Washington DC. We take road and travel to work data from the US census and use this as the basis for our model.
The road data acts as a basis for our agents (red) to move from their homes (areas shaded green) to Tyson’s corner and the census data provides us with the number of agents who travel to the area on a daily basis. The agents attempt to find the shortest path from their home to the destination with preferential attachment to highways and freeways over smaller country roads. By running the model, cars start at homes and travel towards Tyson’s Corner and as more cars join certain sections of roads, traffic jams start to form (as speed is a function of the number of cars on a specific section of road). For example in the movie above, individual cars can be distinguished when they are not clustered but when traffic density increases, larger clusters develop.
The Business Assessment Model (BAM) evaluates trajectories of People, Performance and Planet (3P) of a single company as well as of the whole market system as a function of business decision of actors, exogenous events in the broader socioeconomic environment or both.
BAM computes the cumulative di difference between predictions of the perturbed and original 3P trajectories in order to conduct analysis of decisions within the medium run planning horizon.
Figure: Integrated 3P trajectories comparing system-level dynamics of both Baseline and Variant scenarios
Download self contained, ready to run version of the BAM simulation used in the "Comprehensive Assessment of Businesss Decisions" working paper (zip file) and associated presentation (ppt file) on running and interpreting outputs;
This six-day, non-credit, course is a unique opportunity to work with a team of experienced computational social scientists to explore and understand the application of new interdisciplinary approaches to modeling and making decisions involving the operations of social systems. Participants will imerse themselves in an intensive tour of the field of Computational Social Science, a broad set of efforts that seek to explain and predict how large-scale human systems from organizations to urban systems, from economies to society as a whole, evolve, react to stresses and stimuli, and cooperate and compete. Participants will hear presentations from experts in the field and engage in intensive dialoguing, demonstrations, and policy scenarios.
Computational Social Science (CSS) is a relatively new interdisciplinary science in which social science questions are investigated with modern computational tools. Computational social scientists investigate complex social phenomenon such as economic markets, traffic control, and political systems by simulating the interactions of the many actors in such systems, on computers. They hope to gain insights which will lead to better management of the behavior of the larger social systems, i.e., prevention of market crashes, smoothed traffic flow, or maintenance of political stability. The intractability of many social problems calls for the new approaches provided by computational social science.
CSS is a highly interdisciplinary field that requires teams to plan and complete projects, be they undertaken by government, industry, or non-profit entities. Project managers of such teams, overseeing all elements of project design and execution, tend to hold PhDs. The MAIS concentration will train students to be members of these project teams, able to meaningfully contribute to background research and to project design, execution, and communication.
Prior background should include a bachelor’s degree in one of the social sciences, in computer science, in engineering, or in a relevant discipline, as well as undergraduate courses in these and related areas. Bachelor’s degrees in other areas are also eligible, but the student may be required to take additional courses in social science, mathematics, or computer science as prerequisites to admission.
This concentration will be available in fall 2011.
Concentration Requirements (Catalog Year 2010-2011)
Six core courses (18 credits)
Three required courses (9 credits): CSS 600, 605, 610
The required CSS courses provide an understanding of the conceptual, technical, and practical foundations of computational social science.
The electives provide an understanding of the technical foundations and current work in at least two subfields of computational social science.
One research courses (3 credits) chosen from: CSS 796, 898, 899
The research course provides students with exposure to the most current ongoing research in the field and allows them to further develop their computational research expertise.
Three-four elective courses (9-12 credits)
The electives allow students to acquire a substantive specialization as well as additional training in social and computational science. Because of the broad spectrum of social science phenomena, methodologies, and student backgrounds, there is a large pool of potential courses. Electives may include any Mason master’s-level course in computational social science, social science, computer science, statistics, or other quantitative methods such as data visualization, information technology, and geographic information science. Electives should be selected in conjunction with and approval of the student’s advisor and the Director of CSS Graduate Studies. If the student does not have prior coursework in multivariate statistical analysis, the electives should include at least one such course relevant for the student’s chosen specialization.
Students who elect to do a 5-credit project or a thesis take 9 elective credits. Students who do a 2-credit project take 12 credits.
Proposal (1 credit): MAIS 797
Project (2-5 credits) or thesis (5 credits): MAIS 798 or MAIS 799.
As many of us know more people are now living in cities than ever before. Over half (3.3 billion people) of the world’s population are located in urban areas and this proportion is predicted to increase to over 75 percent by the year 2100 (United Nations, 2007). What I find most interesting is that while cities are growing they are also constantly changing.
The reason I bring this up are twofold. First I recently stumbled upon a site from the NY Times entitled "Mapping America: Every City, Every Block" which shows data from the US Census Bureau's American Community Survey from 2005 to 2009. What is interesting is that there are several maps showing the distribution of racial and ethnic groups throughout the US along with the percentage of foreign born and how neighborhood have changed over time (similar work is also being done at CASA for the UK's population). Such data could come in handy for residential segregation models.
Secondly, as cities grow, the question is where will such growth occur or in some instances where will urban areas decline (e.g. urban shrinkage). Recently we started using GeoMason to explore urban growth using the SLEUTH model as our basis. In the movie below we explore a simple growth scenario around Santa Fe, New Mexico. The model, like the original has the 5 growth coefficients (Dispersion, Breed Spread, Slope, Road) which affect how the growth rules are applied. We also implement the same growth rules (that of Spontaneous Growth, New Spreading Centers, Edge Growth,Road-Influenced Growth) as described on the SLEUTH website. The only thing missing from the model is the self modification procedures during the run time of the model. Data for the model comes a variety of sources including the National Map, the New Mexico Atlas and ESRI.
Over the last few months I have been experimenting with GeoMason in terms of using both raster and vector data to build spatially explicit agent-based models. Overall, I have been impressed by the speed of which such models run at and the example models that come with it GeoMason are really useful.
Below are some of the test models we have developed for a course on entitled GIS and Agent-based modeling which was taught in the Department of Computational Social Science at GMU.
The first model is based on importing a geotiff of the National Elevation Dataset at 1 arc second from the National Map of Crater Lake, Oregon. After which water is added (which could be considered loosely as agents) and flows from high to low elevations, if the water can not flow over the surface it pools. As the movie below shows, over time, Crater Lake slowly fills up until the water breaches the caldera rim which allows the water to flow out. This model was inspired by the NetLogo Grand Canyon Model.
Moving towards vector data, below is a simple example of using GeoMason for a spatially explicit Schelling style of model of segregation where neighborhoods are not based on a regular square lattice but on irregular cells and in this case stylized on Washington DC. The model uses polygon boundaries of census tracks from the 2000 US census. In the model, each census track knows which other census tracts are its neighbors and only one agent occupies such a track. As in the normal Schelling styles of models, agents want to be located in neighborhoods where a certain percentage of their neighbors are of the same type. If an agent is dissatisfied with its neighborhood it moves to a new location. This movement causes more movement and the result is that areas become segregated.
The next movie, highlights a variant of the Schelling model. Whereby we take aggregate data and use this to populate the model with agents. The reason for this is that much of the data we have comes at an aggregate level and often in some sort of vector representation of space such as census data. However, if we want to model the individuals or groups of individuals we can take this data along with other information and use it as a basis of spatially explicit agent-based model. Basically, this model extends the polygon model above and creates agents based on attribute data held within the shapefile. While the model is highly stylized it demonstrates how data can be read into the model, how to create agents based on this data and how to link points (agents) to polygons along with some basic geographical operations (such as union, point in polygon, buffer).
To download GeoMason click here (note there a number of model examples here). For more information on GeoMason there is a technical report:
Sullivan, K., Coletti, M. and Luke, S. (2010), GeoMason: GeoSpatial Support for MASON, Department of Computer Science, George Mason University, Technical Report Series, Fairfax, VA.
Also when I have time I will provide more details on the models above and others using GeoMason.
A short note to say that the videos and presentations from the Multi-disciplinary approach of complexity, networks, geosimulations workshop held between 9th - 11th June 2010 at the University of Lausanne are available online and for download: http://www.unil.ch/citadyne/page78227.html. Presentations include those from Prof. Peter Allen entitled "The Complexity of Structure, Strategy and Decision Making" and Prof. Itzhak Benenson entitled "Geo-simulations of urban phenomena"